Chapter_3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.292652Z",
     "start_time": "2021-03-22T14:56:50.045676Z"
    },
    "tags": [
     "remove_cell"
    ]
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "\n",
    "from tqdm.autonotebook import tqdm\n",
    "\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import imshow\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "import time\n",
    "\n",
    "from idlmam import train_simple_network, set_seed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.300000Z",
     "start_time": "2021-03-22T14:56:51.294314Z"
    },
    "tags": [
     "remove_cell"
    ]
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from IPython.display import set_matplotlib_formats\n",
    "set_matplotlib_formats('png', 'pdf')\n",
    "\n",
    "def set_seed(seed):\n",
    "    torch.manual_seed(seed)\n",
    "    np.random.seed(seed)\n",
    "\n",
    "torch.backends.cudnn.deterministic=True\n",
    "set_seed(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.351972Z",
     "start_time": "2021-03-22T14:56:51.309417Z"
    }
   },
   "outputs": [],
   "source": [
    "import torchvision \n",
    "from torchvision import transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.379280Z",
     "start_time": "2021-03-22T14:56:51.354271Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PIL.Image.Image"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mnist_data_train = torchvision.datasets.MNIST(\"./data\", train=True, download=True)\n",
    "mnist_data_test = torchvision.datasets.MNIST(\"./data\", train=False, download=True)\n",
    "x_example, y_example = mnist_data_train[0]\n",
    "type(x_example)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.408830Z",
     "start_time": "2021-03-22T14:56:51.380844Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 28, 28])\n"
     ]
    }
   ],
   "source": [
    "mnist_data_train = torchvision.datasets.MNIST(\"./data\", train=True, download=True, transform=transforms.ToTensor())\n",
    "mnist_data_test = torchvision.datasets.MNIST(\"./data\", train=False, download=True, transform=transforms.ToTensor())\n",
    "x_example, y_example = mnist_data_train[0]\n",
    "print(x_example.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.595917Z",
     "start_time": "2021-03-22T14:56:51.409942Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6a1fea3090>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "imshow(x_example[0,:], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.601538Z",
     "start_time": "2021-03-22T14:56:51.597195Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([3, 28, 28])\n"
     ]
    }
   ],
   "source": [
    "x_as_color = torch.stack([x_example[0,:], x_example[0,:], x_example[0,:]], dim=0)\n",
    "print(x_as_color.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.764490Z",
     "start_time": "2021-03-22T14:56:51.602584Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6b681c60d0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjEwIDAgb2JqCjw8IC9Bbm5vdHMgWyBdIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDI1MS41NTg3NSAyNDguNTExODc1IF0gL1BhcmVudCAyIDAgUiAvUmVzb3VyY2VzIDggMCBSCi9UeXBlIC9QYWdlID4+CmVuZG9iago5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTEgMCBSID4+CnN0cmVhbQp4nI2VzW7bMBCE73yKPbaXFXe5/NljjLRGe3NroA/gOmkNO0UaoHn9rpTaIlU5ysGQNKbn85BDiuDguhuC+yfwcLDPMxCsobvd//m5239Zr2D35LzpJ8eRMMaSoz0dqyeWgpGovz3a0Obxh3MPztztJ2szvneOE+rwJQcsebgzaxEMsVWPtSqE4ew5OtSqke7cI8zYM2UUOV9+7+EbPEB3w31mssxkmf1/mc2KCvTJ++uM7e4E3SeC21+wcRt4PDt6C9u7eiz/fE1xwR49JZ+b2JXqMZxju5XN2LNbbaH7SEAetneOMyYmJU3CCoqcC3GE7Xf3zr+H7QE+bN1AdEkxBU9ZG1KlLpBSQolelHwmaUmxJZEvNnS6lJW6QCLP6FWDDWgwNElEkpFLDkwtqJKXSOJROavY7JU0oU1TlYREsWhoaZW8RMuKOaZIucRMLY0n2ZgFVcT+WLsfKnmpGWTbrURflA03oU2yVS22vcDqrdkGCyhT9UqyYFt+WDAOwQrFJDxknOnhhUVaMHlvYWpWpS6xSMW2STDOsHozTRxZ0ab+MokX1qjOs6xINiii5vkGjv62GioheGn8R/UVfzt+aChgH2S2ehdMYeuaWLVryiheh+RiB+JQu/6YmC3cmSHBdk+iRDVjFK8zgqK8lI0kTGrG8PnlNTIcfu2BOn/wz57l7uvsG+F07Y3Qj3/7a6UZPdq85r5xfwFEfnPICmVuZHN0cmVhbQplbmRvYmoKMTEgMCBvYmoKNTQxCmVuZG9iagoxNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0NyA+PgpzdHJlYW0KeJxNUbttRDEM698UXOAA62t5ngtSXfZvQ8kIkMIgoS8ppyUW9sZLDOEHWw++5JFVQ38ePzHsMyw9yeTUP+a5yVQUvhWqm5hQF2Lh/WgEvBZ0LyIrygffj2UMc8734KMQl2AmNGCsb0kmF9W8M2TCiaGOw0GbVBh3TRQsrhXNM8jtVjeyOrMgbHglE+LGAEQE2ReQzWCjjLGVkMVyHqgKkgVaYNfpG1GLgiuU1gl0otbEuszgq+f2djdDL/LgqLp4fQzrS7DC6KV7LHyuQh/M9Ew7d0kjvfCmExFmDwVSmZ2RlTo9Yn23QP+fZSv4+8nP8/0LFShcKgplbmRzdHJlYW0KZW5kb2JqCjE4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggODAgPj4Kc3RyZWFtCnicRYy7DcAwCER7pmAEfiZmnyiVs38bIErccE+6e7g6EjJT3mGGhwSeDCyGU/EGmaNgNbhGUo2d7KOwbl91geZ6U6v19wcqT3Z2cT3Nyxn0CmVuZHN0cmVhbQplbmRvYmoKMTkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDggPj4Kc3RyZWFtCnicLVE5kgNBCMvnFXpCc9PvscuR9//pCsoBg4ZDIDotcVDGTxCWK97yyFW04e+ZGMF3waHfynUbFjkQFUjSGFRNqF28Hr0HdhxmAvOkNSyDGesDP2MKN3pxeEzG2e11GTUEe9drT2ZQMisXccnEBVN12MiZw0+mjAvtXM8NyLkR1mUYpJuVxoyEI00hUkih6iapM0GQBKOrUaONHMV+6csjnWFVI2oM+1xL29dzE84aNDsWqzw5pUdXnMvJxQsrB/28zcBFVBqrPBAScL/bQ/2c7OQ33tK5s8X0+F5zsrwwFVjx5rUbkE21+Dcv4vg94+v5/AOopVsWCmVuZHN0cmVhbQplbmRvYmoKMjAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTAgPj4Kc3RyZWFtCnicNVDLDUMxCLtnChaoFAKBZJ5WvXX/a23QO2ER/0JYyJQIeanJzinpSz46TA+2Lr+xIgutdSXsypognivvoZmysdHY4mBwGiZegBY3YOhpjRo1dOGCpi6VQoHFJfCZfHV76L5PGXhqGXJ2BBFDyWAJaroWTVi0PJ+QTgHi/37D7i3koZLzyp4b+Ruc7fA7s27hJ2p2ItFyFTLUszTHGAgTRR48eUWmcOKz1nfVNBLUZgtOlgGuTj+MDgBgIl5ZgOyuRDlL0o6ln2+8x/cPQABTtAplbmRzdHJlYW0KZW5kb2JqCjE1IDAgb2JqCjw8IC9CYXNlRm9udCAvRGVqYVZ1U2FucyAvQ2hhclByb2NzIDE2IDAgUgovRW5jb2RpbmcgPDwgL0RpZmZlcmVuY2VzIFsgNDggL3plcm8gL29uZSAvdHdvIDUzIC9maXZlIF0gL1R5cGUgL0VuY29kaW5nID4+Ci9GaXJzdENoYXIgMCAvRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250RGVzY3JpcHRvciAxNCAwIFIKL0ZvbnRNYXRyaXggWyAwLjAwMSAwIDAgMC4wMDEgMCAwIF0gL0xhc3RDaGFyIDI1NSAvTmFtZSAvRGVqYVZ1U2FucwovU3VidHlwZSAvVHlwZTMgL1R5cGUgL0ZvbnQgL1dpZHRocyAxMyAwIFIgPj4KZW5kb2JqCjE0IDAgb2JqCjw8IC9Bc2NlbnQgOTI5IC9DYXBIZWlnaHQgMCAvRGVzY2VudCAtMjM2IC9GbGFncyAzMgovRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250TmFtZSAvRGVqYVZ1U2FucyAvSXRhbGljQW5nbGUgMAovTWF4V2lkdGggMTM0MiAvU3RlbVYgMCAvVHlwZSAvRm9udERlc2NyaXB0b3IgL1hIZWlnaHQgMCA+PgplbmRvYmoKMTMgMCBvYmoKWyA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMAo2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDMxOCA0MDEgNDYwIDgzOCA2MzYKOTUwIDc4MCAyNzUgMzkwIDM5MCA1MDAgODM4IDMxOCAzNjEgMzE4IDMzNyA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2CjYzNiA2MzYgMzM3IDMzNyA4MzggODM4IDgzOCA1MzEgMTAwMCA2ODQgNjg2IDY5OCA3NzAgNjMyIDU3NSA3NzUgNzUyIDI5NQoyOTUgNjU2IDU1NyA4NjMgNzQ4IDc4NyA2MDMgNzg3IDY5NSA2MzUgNjExIDczMiA2ODQgOTg5IDY4NSA2MTEgNjg1IDM5MCAzMzcKMzkwIDgzOCA1MDAgNTAwIDYxMyA2MzUgNTUwIDYzNSA2MTUgMzUyIDYzNSA2MzQgMjc4IDI3OCA1NzkgMjc4IDk3NCA2MzQgNjEyCjYzNSA2MzUgNDExIDUyMSAzOTIgNjM0IDU5MiA4MTggNTkyIDU5MiA1MjUgNjM2IDMzNyA2MzYgODM4IDYwMCA2MzYgNjAwIDMxOAozNTIgNTE4IDEwMDAgNTAwIDUwMCA1MDAgMTM0MiA2MzUgNDAwIDEwNzAgNjAwIDY4NSA2MDAgNjAwIDMxOCAzMTggNTE4IDUxOAo1OTAgNTAwIDEwMDAgNTAwIDEwMDAgNTIxIDQwMCAxMDIzIDYwMCA1MjUgNjExIDMxOCA0MDEgNjM2IDYzNiA2MzYgNjM2IDMzNwo1MDAgNTAwIDEwMDAgNDcxIDYxMiA4MzggMzYxIDEwMDAgNTAwIDUwMCA4MzggNDAxIDQwMSA1MDAgNjM2IDYzNiAzMTggNTAwCjQwMSA0NzEgNjEyIDk2OSA5NjkgOTY5IDUzMSA2ODQgNjg0IDY4NCA2ODQgNjg0IDY4NCA5NzQgNjk4IDYzMiA2MzIgNjMyIDYzMgoyOTUgMjk1IDI5NSAyOTUgNzc1IDc0OCA3ODcgNzg3IDc4NyA3ODcgNzg3IDgzOCA3ODcgNzMyIDczMiA3MzIgNzMyIDYxMSA2MDUKNjMwIDYxMyA2MTMgNjEzIDYxMyA2MTMgNjEzIDk4MiA1NTAgNjE1IDYxNSA2MTUgNjE1IDI3OCAyNzggMjc4IDI3OCA2MTIgNjM0CjYxMiA2MTIgNjEyIDYxMiA2MTIgODM4IDYxMiA2MzQgNjM0IDYzNCA2MzQgNTkyIDYzNSA1OTIgXQplbmRvYmoKMTYgMCBvYmoKPDwgL2ZpdmUgMTcgMCBSIC9vbmUgMTggMCBSIC90d28gMTkgMCBSIC96ZXJvIDIwIDAgUiA+PgplbmRvYmoKMyAwIG9iago8PCAvRjEgMTUgMCBSID4+CmVuZG9iago0IDAgb2JqCjw8IC9BMSA8PCAvQ0EgMCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+Ci9BMiA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+ID4+CmVuZG9iago1IDAgb2JqCjw8ID4+CmVuZG9iago2IDAgb2JqCjw8ID4+CmVuZG9iago3IDAgb2JqCjw8IC9JMSAxMiAwIFIgPj4KZW5kb2JqCjEyIDAgb2JqCjw8IC9CaXRzUGVyQ29tcG9uZW50IDggL0NvbG9yU3BhY2UgL0RldmljZVJHQgovRGVjb2RlUGFybXMgPDwgL0NvbG9ycyAzIC9Db2x1bW5zIDIxOCAvUHJlZGljdG9yIDEwID4+Ci9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9IZWlnaHQgMjE4IC9MZW5ndGggMjEgMCBSIC9TdWJ0eXBlIC9JbWFnZQovVHlwZSAvWE9iamVjdCAvV2lkdGggMjE4ID4+CnN0cmVhbQp4nO3dTYjNexzH8XHJgjDNhlLSWIyaiAVKStIkxWJiQ9lgN2Q1GzsLUmYWyGJWykK2zIryuFBTk4cN2XvYMWjyEOPu7/me2znumeszvF7LT//O/dV991v8O+N0dQEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAABd8371AULNnz+/3JctW9aRzz927FjjuGjRovLhvr6+ch8aGir3kZGRxvHAgQPlw58/fy73s2fPlvupU6fKvSP+mr2PhnbJkSByJIgcCSJHgsiRIHIkyIJffYCft2rVqnJfuHBhuW/durXct23b1jh2d3eXD+/bt6+lw3XUy5cvy/3ChQvlPjg42Dh+/PixfPjp06flfv/+/dZO10luR4LIkSByJIgcCSJHgsiRIHPgC2YbN24s99u3b5d7p74D9v+bmZkp98OHD5f79PR06x/++vXrcn/37l25v3jxovUP7xS3I0HkSBA5EkSOBJEjQeRIEDkSZA68d+zp6Sn3iYmJcu/t7Z3N49SaHWZqaqrcd+zY0Th+/fq1fHjuvkltl9uRIHIkiBwJIkeCyJEgciSIHAkyB/6w9e3bt+U+PDxc7nv27Cn3x48fl3uzPw8tPXnypNwHBgbKvdlXEvv7+xvHEydOtH6S35LbkSByJIgcCSJHgsiRIHIkiBwJMge+79iupUuXlnuzf1FubGyscTxy5Ej58KFDh8r96tWrrZ2Of+N2JIgcCSJHgsiRIHIkiBwJIkeCzIHvO7brw4cPbT3//v371h8+evRouV+7dq3cm/2TjZTcjgSRI0HkSBA5EkSOBJEjQX7DL5i1a/HixY3j+Ph4+fD27dvLfffu3eV+69atnz7YH8jtSBA5EkSOBJEjQeRIEDkSRI4E8d6xtmbNmnJ/9OhRuTf7wY67d++W++TkZON46dKl8uEfP36U++/H7UgQORJEjgSRI0HkSBA5EkSOBPHesT2Dg4Plfvny5XJfsmRJ6x9+8uTJcr9y5Uq5v3nzpvUPnxPcjgSRI0HkSBA5EkSOBJEjQeRIEO8dO2PdunXlPjo6Wu47d+5s/cPLnxrp6uo6ffp0ub969ar1D4/idiSIHAkiR4LIkSByJIgcCSJHgnjvOLu6u7vLfe/evY1jsy9NzptX/2+6c+dOuQ8MDLR0uDxuR4LIkSByJIgcCSJHgsiRIF70BPny5Uu5L1hQ/7Dut2/fyn3Xrl2N47179372XP8ftyNB5EgQORJEjgSRI0HkSBA5EqR+oUW71q9fX+779+8v902bNjWOzd4vNvPs2bNyf/DgQVufk8PtSBA5EkSOBJEjQeRIEDkSRI4E8d6x1tfXV+7Hjx8v92Y/8LFixYr/fpjv37+Xe7Mf8piZmfnv/9Ffwu1IEDkSRI4EkSNB5EgQORJEjgT5g/7OutkrwIMHDzaOQ0ND5cOrV6/u4JH+YXJystyb/WDHjRs3Zu8wv4TbkSByJIgcCSJHgsiRIHIkyBx+0bN8+fJy7+/vL/eLFy+W+9q1azt2pgYTExPlfu7cucbx+vXr5cNz9wtj7XI7EkSOBJEjQeRIEDkSRI4EkSNBst479vT0NI5jY2Plwxs2bCj33t7eDh7pHx4+fFjuo6Oj5X7z5s1y//TpU8fO9BtxOxJEjgSRI0HkSBA5EkSOBJEjQWb3veOWLVvKfXh4uNw3b97cOK5cubKTZ2pQvgI8f/58+fCZM2fKfXp6upNn+lO5HQkiR4LIkSByJIgcCSJHgsiRILP7Qx7Nft6i2d6W58+fl/v4+Hi5N/s5jJGRkcZxamrqZ8/Fz3M7EkSOBJEjQeRIEDkSRI4EkSMAAAAAAAAAAAAAAAAAAPCn+RtDReR6CmVuZHN0cmVhbQplbmRvYmoKMjEgMCBvYmoKMTM3OAplbmRvYmoKMiAwIG9iago8PCAvQ291bnQgMSAvS2lkcyBbIDEwIDAgUiBdIC9UeXBlIC9QYWdlcyA+PgplbmRvYmoKMjIgMCBvYmoKPDwgL0NyZWF0aW9uRGF0ZSAoRDoyMDIxMDMyMjEwNTY1MS0wNCcwMCcpCi9DcmVhdG9yIChNYXRwbG90bGliIHYzLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZykKL1Byb2R1Y2VyIChNYXRwbG90bGliIHBkZiBiYWNrZW5kIHYzLjMuMikgPj4KZW5kb2JqCnhyZWYKMCAyMwowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDAwMTYgMDAwMDAgbiAKMDAwMDAwNTU4OSAwMDAwMCBuIAowMDAwMDAzNzU2IDAwMDAwIG4gCjAwMDAwMDM3ODggMDAwMDAgbiAKMDAwMDAwMzg4NyAwMDAwMCBuIAowMDAwMDAzOTA4IDAwMDAwIG4gCjAwMDAwMDM5MjkgMDAwMDAgbiAKMDAwMDAwMDA2NSAwMDAwMCBuIAowMDAwMDAwMzk4IDAwMDAwIG4gCjAwMDAwMDAyMDggMDAwMDAgbiAKMDAwMDAwMTAxNCAwMDAwMCBuIAowMDAwMDAzOTYxIDAwMDAwIG4gCjAwMDAwMDI2MzEgMDAwMDAgbiAKMDAwMDAwMjQzMSAwMDAwMCBuIAowMDAwMDAyMTEwIDAwMDAwIG4gCjAwMDAwMDM2ODQgMDAwMDAgbiAKMDAwMDAwMTAzNCAwMDAwMCBuIAowMDAwMDAxMzU0IDAwMDAwIG4gCjAwMDAwMDE1MDYgMDAwMDAgbiAKMDAwMDAwMTgyNyAwMDAwMCBuIAowMDAwMDA1NTY4IDAwMDAwIG4gCjAwMDAwMDU2NDkgMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyAyMiAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgMjMgPj4Kc3RhcnR4cmVmCjU4MDYKJSVFT0YK\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "imshow(x_as_color.permute(1,2,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:51.929352Z",
     "start_time": "2021-03-22T14:56:51.765727Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6a1fc8b810>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD4CAYAAAAq5pAIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAMqElEQVR4nO3dbawcZRnG8euSlw9WlFZCbbAKGOILvhRTqxGiGAMi0ZTGl1ATUkP18AEMJIbY4AcxBkNUNGqM4RiQYhRiAoRqjEAqUo0J4YAFChWKpGrpSStBpBgVKbcfdmqO7dl9ztmZ2dme+/9LTnZ37p2dO5tendl9ZvZxRAjAwveyrhsAMBqEHUiCsANJEHYgCcIOJHHkKDdmm6/+gZZFhGdbXmvPbvsc24/ZfsL2hjqvBaBdHnac3fYRkh6XdJakXZLuk7Q2Ih4dsA57dqBlbezZV0l6IiKejIgXJN0saXWN1wPQojphP0HSX2Y83lUt+z+2J2xP2Z6qsS0ANdX5gm62Q4VDDtMjYlLSpMRhPNClOnv2XZKWz3j8Wkm767UDoC11wn6fpFNsn2T7aEnnS9rUTFsAmjb0YXxEvGj7Ekl3SDpC0vUR8UhjnQFo1NBDb0NtjM/sQOtaOakGwOGDsANJEHYgCcIOJEHYgSQIO5AEYQeSIOxAEoQdSIKwA0kQdiAJwg4kQdiBJAg7kARhB5Ig7EAShB1IgrADSRB2IAnCDiRB2IEkCDuQBGEHkiDsQBKEHUiCsANJEHYgCcIOJEHYgSSGnrIZh4kjCvVXtbz9SwbUXl5Y942F+sWF+jcG1NYW1v1XoX51of7lQr0DtcJue6ekfZL2S3oxIlY20RSA5jWxZ/9ARDzdwOsAaBGf2YEk6oY9JN1p+37bE7M9wfaE7SnbUzW3BaCGuofxp0fEbtvHS7rL9h8iYsvMJ0TEpKRJSbIdNbcHYEi19uwRsbu63SvpNkmrmmgKQPOGDrvtRbaPOXBf0tmStjXVGIBm1TmMXyrpNtsHXucnEfHLRrpaaF5XqB9dqL+3UD9jQO3YwrofK9S7tKtQ/06hvmZAbV9h3QcL9XsK9TE0dNgj4klJ72iwFwAtYugNSIKwA0kQdiAJwg4kQdiBJBwxupPaFuwZdKcV6psL9bYvMx1XLxXqFxbq/6ix7d2F+t8K9cdqbLtlEeHZlrNnB5Ig7EAShB1IgrADSRB2IAnCDiRB2IEkGGdvwpJC/d5C/eSmGmlBqfdnC/UPDKi9UFg36/kHNTHODiRH2IEkCDuQBGEHkiDsQBKEHUiCsANJMGVzE54p1C8v1D9SqP++UC/9pPIgWwv1swr10jXlpw6oXVpYF41izw4kQdiBJAg7kARhB5Ig7EAShB1IgrADSXA9+zh4ZaFeml742gG19YV1LyjUf1KoY+wMfT277ett77W9bcayJbbvsr2jul3cZLMAmjeXw/gbJJ1z0LINkjZHxCnqzXeyoeG+ADSsGPaI2KJDTwhdLWljdX+jpPOabQtA04Y9N35pRExLUkRM2z6+3xNtT0iaGHI7ABrS+oUwETEpaVLiCzqgS8MOve2xvUySqtu9zbUEoA3Dhn2TpHXV/XWSbm+mHQBtKR7G275J0pmSjrO9S9KXJF0t6ae210v6s6RPtNnkgvdczfX/XmPdzxTqNxfqpTnWMTaKYY+ItX1KH2y4FwAt4nRZIAnCDiRB2IEkCDuQBGEHkuAS14Vg0YDazwrrvr9Q/3ChfmehjpFjymYgOcIOJEHYgSQIO5AEYQeSIOxAEoQdSIJx9oXuDYX6A4X6s4X63YX61IDa9wrr8q9lKIyzA8kRdiAJwg4kQdiBJAg7kARhB5Ig7EASjLNnt6ZQ/2GhfkyNbV9RqN9YqE/X2PYCxjg7kBxhB5Ig7EAShB1IgrADSRB2IAnCDiTBODsGe1uhfk2hXmeu32sL9asK9adqbPswNvQ4u+3rbe+1vW3GsittP2V7a/V3bpPNAmjeXA7jb5B0zizLvxURK6q/XzTbFoCmFcMeEVskPTOCXgC0qM4XdJfYfqg6zF/c70m2J2xP2R70a2QAWjZs2L+v3k8ZrlDvcoS+X9NExGRErIyIlUNuC0ADhgp7ROyJiP0R8ZKkH0ha1WxbAJo2VNhtL5vxcI2kbf2eC2A8FMfZbd8k6UxJx0naI+lL1eMV6v2y905JF0VE8epixtkXoGML9Y8OqJWulZ91tHiGXxXqZxXqC1S/cfYj57Di2lkWX1e7IwAjxemyQBKEHUiCsANJEHYgCcIOJMElrujOvwv10ljRi4X6hwbUfl1Y9zDGT0kDyRF2IAnCDiRB2IEkCDuQBGEHkiDsQBLFq96Q3NsL9Y8X6u8aUKv7r+/RQn1LzddfYNizA0kQdiAJwg4kQdiBJAg7kARhB5Ig7EASjLMvdG8s1D9XqK8p1F8zj17ma3+hXvrx8peaamRhYM8OJEHYgSQIO5AEYQeSIOxAEoQdSIKwA0kwzn44KI1lf2pA7eLCuifOr5VGTRXqVxXqm5pqJIfint32ctt3295u+xHbl1bLl9i+y/aO6nZx++0CGNZcDuNflPT5iHizpPdIutj2WyRtkLQ5Ik6RtLl6DGBMFcMeEdMR8UB1f5+k7ZJOkLRa0sbqaRslnddSjwAaMK/P7LZPlHSapHslLY2Iaan3H4Lt4/usMyFpomafAGqac9htv0LSLZIui4jn7FnnjjtERExKmqxeg4kdgY7MaejN9lHqBf3HEXFrtXiP7WVVfZmkve20CKAJxT27e7vw6yRtj4hvzihtkrRO0tXV7e2tdLgQLC3UTy3Uv1uov2kevTTt3kL96wNqpX8xXKLaqLkcxp8u6QJJD9veWi27Qr2Q/9T2ekl/lvSJVjoE0Ihi2CPit5L6fUD/YLPtAGgLp8sCSRB2IAnCDiRB2IEkCDuQBJe4ztWSAbVrC+uuKNRPnl8rjfpdoX5NoX5Hof7PefSCVrFnB5Ig7EAShB1IgrADSRB2IAnCDiRB2IEk8oyzv7tQv7xQXzWgdsI8e2naoLHsbxfW/Wqh/o959oKxxZ4dSIKwA0kQdiAJwg4kQdiBJAg7kARhB5LIM86+pma9ju2F+s8K9f2F+jcG1J4trIs02LMDSRB2IAnCDiRB2IEkCDuQBGEHkiDsQBKOiMFPsJdLulHSa9SbMXsyIr5t+0pJn5X01+qpV0TELwqvNXhjAGqLiFlnXZ5L2JdJWhYRD9g+RtL9ks6T9ElJz0fEoFM6Dn4twg60rF/Y5zI/+7Sk6er+Ptvb1f1vswCYp3l9Zrd9oqTTJN1bLbrE9kO2r7e9uM86E7anbE/VaxVAHcXD+P890X6FpHskXRURt9peKulpSSHpK+od6l9YeA0O44GWDf2ZXZJsHyXp55LuiIhvzlI/UdLPI+Kthdch7EDL+oW9eBhv25Kuk7R9ZtCrL+4OWCNpW90mAbRnLt/GnyHpN5IeVm/oTZKukLRWvcmIQ9JOSRdVX+YNei327EDLah3GN4WwA+0b+jAewMJA2IEkCDuQBGEHkiDsQBKEHUiCsANJEHYgCcIOJEHYgSQIO5AEYQeSIOxAEoQdSGLUUzY/LelPMx4fVy0bR+Pa27j2JdHbsJrs7fX9CiO9nv2QjdtTEbGyswYGGNfexrUvid6GNareOIwHkiDsQBJdh32y4+0PMq69jWtfEr0NayS9dfqZHcDodL1nBzAihB1IopOw2z7H9mO2n7C9oYse+rG90/bDtrd2PT9dNYfeXtvbZixbYvsu2zuq21nn2OuotyttP1W9d1ttn9tRb8tt3217u+1HbF9aLe/0vRvQ10jet5F/Zrd9hKTHJZ0laZek+yStjYhHR9pIH7Z3SloZEZ2fgGH7fZKel3Tjgam1bH9N0jMRcXX1H+XiiPjCmPR2peY5jXdLvfWbZvzT6vC9a3L682F0sWdfJemJiHgyIl6QdLOk1R30MfYiYoukZw5avFrSxur+RvX+sYxcn97GQkRMR8QD1f19kg5MM97pezegr5HoIuwnSPrLjMe7NF7zvYekO23fb3ui62ZmsfTANFvV7fEd93Ow4jTeo3TQNONj894NM/15XV2EfbapacZp/O/0iHinpA9Lurg6XMXcfF/SG9SbA3Ba0jVdNlNNM36LpMsi4rkue5lplr5G8r51EfZdkpbPePxaSbs76GNWEbG7ut0r6Tb1PnaMkz0HZtCtbvd23M//RMSeiNgfES9J+oE6fO+qacZvkfTjiLi1Wtz5ezdbX6N637oI+32STrF9ku2jJZ0vaVMHfRzC9qLqixPZXiTpbI3fVNSbJK2r7q+TdHuHvfyfcZnGu9804+r4vet8+vOIGPmfpHPV+0b+j5K+2EUPffo6WdKD1d8jXfcm6Sb1Duv+o94R0XpJr5a0WdKO6nbJGPX2I/Wm9n5IvWAt66i3M9T7aPiQpK3V37ldv3cD+hrJ+8bpskASnEEHJEHYgSQIO5AEYQeSIOxAEoQdSIKwA0n8F7r4z1b/TCnaAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_as_color = torch.stack([x_example[0,:], x_example[0,:], x_example[0,:]])\n",
    "x_as_color[0,:] = 0 #No Red\n",
    "#Leaving green alone\n",
    "x_as_color[2,:] = 0 #No Blue\n",
    "imshow(x_as_color.permute(1,2,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.092769Z",
     "start_time": "2021-03-22T14:56:51.930590Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6a1fc00650>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjEwIDAgb2JqCjw8IC9Bbm5vdHMgWyBdIC9Db250ZW50cyA5IDAgUgovR3JvdXAgPDwgL0NTIC9EZXZpY2VSR0IgL1MgL1RyYW5zcGFyZW5jeSAvVHlwZSAvR3JvdXAgPj4KL01lZGlhQm94IFsgMCAwIDI1MS41NTg3NSAyNDguNTExODc1IF0gL1BhcmVudCAyIDAgUiAvUmVzb3VyY2VzIDggMCBSCi9UeXBlIC9QYWdlID4+CmVuZG9iago5IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTEgMCBSID4+CnN0cmVhbQp4nI2VzW7bMBCE73yKPbaXFXe5/NljjLRGe3NroA/gOmkNO0UaoHn9rpTaIlU5ysGQNKbn85BDiuDguhuC+yfwcLDPMxCsobvd//m5239Zr2D35LzpJ8eRMMaSoz0dqyeWgpGovz3a0Obxh3MPztztJ2szvneOE+rwJQcsebgzaxEMsVWPtSqE4ew5OtSqke7cI8zYM2UUOV9+7+EbPEB3w31mssxkmf1/mc2KCvTJ++uM7e4E3SeC21+wcRt4PDt6C9u7eiz/fE1xwR49JZ+b2JXqMZxju5XN2LNbbaH7SEAetneOMyYmJU3CCoqcC3GE7Xf3zr+H7QE+bN1AdEkxBU9ZG1KlLpBSQolelHwmaUmxJZEvNnS6lJW6QCLP6FWDDWgwNElEkpFLDkwtqJKXSOJROavY7JU0oU1TlYREsWhoaZW8RMuKOaZIucRMLY0n2ZgFVcT+WLsfKnmpGWTbrURflA03oU2yVS22vcDqrdkGCyhT9UqyYFt+WDAOwQrFJDxknOnhhUVaMHlvYWpWpS6xSMW2STDOsHozTRxZ0ab+MokX1qjOs6xINiii5vkGjv62GioheGn8R/UVfzt+aChgH2S2ehdMYeuaWLVryiheh+RiB+JQu/6YmC3cmSHBdk+iRDVjFK8zgqK8lI0kTGrG8PnlNTIcfu2BOn/wz57l7uvsG+F07Y3Qj3/7a6UZPdq85r5xfwFEfnPICmVuZHN0cmVhbQplbmRvYmoKMTEgMCBvYmoKNTQxCmVuZG9iagoxNyAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDI0NyA+PgpzdHJlYW0KeJxNUbttRDEM698UXOAA62t5ngtSXfZvQ8kIkMIgoS8ppyUW9sZLDOEHWw++5JFVQ38ePzHsMyw9yeTUP+a5yVQUvhWqm5hQF2Lh/WgEvBZ0LyIrygffj2UMc8734KMQl2AmNGCsb0kmF9W8M2TCiaGOw0GbVBh3TRQsrhXNM8jtVjeyOrMgbHglE+LGAEQE2ReQzWCjjLGVkMVyHqgKkgVaYNfpG1GLgiuU1gl0otbEuszgq+f2djdDL/LgqLp4fQzrS7DC6KV7LHyuQh/M9Ew7d0kjvfCmExFmDwVSmZ2RlTo9Yn23QP+fZSv4+8nP8/0LFShcKgplbmRzdHJlYW0KZW5kb2JqCjE4IDAgb2JqCjw8IC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggODAgPj4Kc3RyZWFtCnicRYy7DcAwCER7pmAEfiZmnyiVs38bIErccE+6e7g6EjJT3mGGhwSeDCyGU/EGmaNgNbhGUo2d7KOwbl91geZ6U6v19wcqT3Z2cT3Nyxn0CmVuZHN0cmVhbQplbmRvYmoKMTkgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyNDggPj4Kc3RyZWFtCnicLVE5kgNBCMvnFXpCc9PvscuR9//pCsoBg4ZDIDotcVDGTxCWK97yyFW04e+ZGMF3waHfynUbFjkQFUjSGFRNqF28Hr0HdhxmAvOkNSyDGesDP2MKN3pxeEzG2e11GTUEe9drT2ZQMisXccnEBVN12MiZw0+mjAvtXM8NyLkR1mUYpJuVxoyEI00hUkih6iapM0GQBKOrUaONHMV+6csjnWFVI2oM+1xL29dzE84aNDsWqzw5pUdXnMvJxQsrB/28zcBFVBqrPBAScL/bQ/2c7OQ33tK5s8X0+F5zsrwwFVjx5rUbkE21+Dcv4vg94+v5/AOopVsWCmVuZHN0cmVhbQplbmRvYmoKMjAgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTAgPj4Kc3RyZWFtCnicNVDLDUMxCLtnChaoFAKBZJ5WvXX/a23QO2ER/0JYyJQIeanJzinpSz46TA+2Lr+xIgutdSXsypognivvoZmysdHY4mBwGiZegBY3YOhpjRo1dOGCpi6VQoHFJfCZfHV76L5PGXhqGXJ2BBFDyWAJaroWTVi0PJ+QTgHi/37D7i3koZLzyp4b+Ruc7fA7s27hJ2p2ItFyFTLUszTHGAgTRR48eUWmcOKz1nfVNBLUZgtOlgGuTj+MDgBgIl5ZgOyuRDlL0o6ln2+8x/cPQABTtAplbmRzdHJlYW0KZW5kb2JqCjE1IDAgb2JqCjw8IC9CYXNlRm9udCAvRGVqYVZ1U2FucyAvQ2hhclByb2NzIDE2IDAgUgovRW5jb2RpbmcgPDwgL0RpZmZlcmVuY2VzIFsgNDggL3plcm8gL29uZSAvdHdvIDUzIC9maXZlIF0gL1R5cGUgL0VuY29kaW5nID4+Ci9GaXJzdENoYXIgMCAvRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250RGVzY3JpcHRvciAxNCAwIFIKL0ZvbnRNYXRyaXggWyAwLjAwMSAwIDAgMC4wMDEgMCAwIF0gL0xhc3RDaGFyIDI1NSAvTmFtZSAvRGVqYVZ1U2FucwovU3VidHlwZSAvVHlwZTMgL1R5cGUgL0ZvbnQgL1dpZHRocyAxMyAwIFIgPj4KZW5kb2JqCjE0IDAgb2JqCjw8IC9Bc2NlbnQgOTI5IC9DYXBIZWlnaHQgMCAvRGVzY2VudCAtMjM2IC9GbGFncyAzMgovRm9udEJCb3ggWyAtMTAyMSAtNDYzIDE3OTQgMTIzMyBdIC9Gb250TmFtZSAvRGVqYVZ1U2FucyAvSXRhbGljQW5nbGUgMAovTWF4V2lkdGggMTM0MiAvU3RlbVYgMCAvVHlwZSAvRm9udERlc2NyaXB0b3IgL1hIZWlnaHQgMCA+PgplbmRvYmoKMTMgMCBvYmoKWyA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMAo2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDYwMCA2MDAgNjAwIDMxOCA0MDEgNDYwIDgzOCA2MzYKOTUwIDc4MCAyNzUgMzkwIDM5MCA1MDAgODM4IDMxOCAzNjEgMzE4IDMzNyA2MzYgNjM2IDYzNiA2MzYgNjM2IDYzNiA2MzYgNjM2CjYzNiA2MzYgMzM3IDMzNyA4MzggODM4IDgzOCA1MzEgMTAwMCA2ODQgNjg2IDY5OCA3NzAgNjMyIDU3NSA3NzUgNzUyIDI5NQoyOTUgNjU2IDU1NyA4NjMgNzQ4IDc4NyA2MDMgNzg3IDY5NSA2MzUgNjExIDczMiA2ODQgOTg5IDY4NSA2MTEgNjg1IDM5MCAzMzcKMzkwIDgzOCA1MDAgNTAwIDYxMyA2MzUgNTUwIDYzNSA2MTUgMzUyIDYzNSA2MzQgMjc4IDI3OCA1NzkgMjc4IDk3NCA2MzQgNjEyCjYzNSA2MzUgNDExIDUyMSAzOTIgNjM0IDU5MiA4MTggNTkyIDU5MiA1MjUgNjM2IDMzNyA2MzYgODM4IDYwMCA2MzYgNjAwIDMxOAozNTIgNTE4IDEwMDAgNTAwIDUwMCA1MDAgMTM0MiA2MzUgNDAwIDEwNzAgNjAwIDY4NSA2MDAgNjAwIDMxOCAzMTggNTE4IDUxOAo1OTAgNTAwIDEwMDAgNTAwIDEwMDAgNTIxIDQwMCAxMDIzIDYwMCA1MjUgNjExIDMxOCA0MDEgNjM2IDYzNiA2MzYgNjM2IDMzNwo1MDAgNTAwIDEwMDAgNDcxIDYxMiA4MzggMzYxIDEwMDAgNTAwIDUwMCA4MzggNDAxIDQwMSA1MDAgNjM2IDYzNiAzMTggNTAwCjQwMSA0NzEgNjEyIDk2OSA5NjkgOTY5IDUzMSA2ODQgNjg0IDY4NCA2ODQgNjg0IDY4NCA5NzQgNjk4IDYzMiA2MzIgNjMyIDYzMgoyOTUgMjk1IDI5NSAyOTUgNzc1IDc0OCA3ODcgNzg3IDc4NyA3ODcgNzg3IDgzOCA3ODcgNzMyIDczMiA3MzIgNzMyIDYxMSA2MDUKNjMwIDYxMyA2MTMgNjEzIDYxMyA2MTMgNjEzIDk4MiA1NTAgNjE1IDYxNSA2MTUgNjE1IDI3OCAyNzggMjc4IDI3OCA2MTIgNjM0CjYxMiA2MTIgNjEyIDYxMiA2MTIgODM4IDYxMiA2MzQgNjM0IDYzNCA2MzQgNTkyIDYzNSA1OTIgXQplbmRvYmoKMTYgMCBvYmoKPDwgL2ZpdmUgMTcgMCBSIC9vbmUgMTggMCBSIC90d28gMTkgMCBSIC96ZXJvIDIwIDAgUiA+PgplbmRvYmoKMyAwIG9iago8PCAvRjEgMTUgMCBSID4+CmVuZG9iago0IDAgb2JqCjw8IC9BMSA8PCAvQ0EgMCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+Ci9BMiA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+ID4+CmVuZG9iago1IDAgb2JqCjw8ID4+CmVuZG9iago2IDAgb2JqCjw8ID4+CmVuZG9iago3IDAgb2JqCjw8IC9JMSAxMiAwIFIgPj4KZW5kb2JqCjEyIDAgb2JqCjw8IC9CaXRzUGVyQ29tcG9uZW50IDggL0NvbG9yU3BhY2UgL0RldmljZVJHQgovRGVjb2RlUGFybXMgPDwgL0NvbG9ycyAzIC9Db2x1bW5zIDIxOCAvUHJlZGljdG9yIDEwID4+Ci9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9IZWlnaHQgMjE4IC9MZW5ndGggMjEgMCBSIC9TdWJ0eXBlIC9JbWFnZQovVHlwZSAvWE9iamVjdCAvV2lkdGggMjE4ID4+CnN0cmVhbQp4nO3dW4iVVQDF8e0FxYoMJTSKbmAGUk6JlhJqmRFplNkFKyWoFNIQCiHCwgil0nlQKZQkSzPsQbxUiEWWEZmoqYRmWYGhDV3MUZssybHHillr8MA5naX9f49rtvt8Z1yzHzbf9+1SAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA4BXWo9wWc7gaYfIoKJ5jBS0w+3+SftXdFyTrW+wKAv1FHBKGOCEIdEYQ6Igh1RBDqiCDsO2qdTN7d/YMGk6/X8ZSzRXiGmaOvyScf0vmcniPahuPKMjn49zJM5s+VL2X+jLmYqmB1RBDqiCDUEUGoI4JQRwShjghCHRGkc43nH2ryniZfefJTX2jyLiYfYvLrVHiOGTx2kPnBCpO7jcoTKjyix+47pvN55rc4ZvBAMffWLXLwDjP5Bh3XFqsjglBHBKGOCEIdEYQ6Igh1RJBab/QMN3kfk4uNnqvM0PdNbu8Bc+R9XVebwa+b/LxKP1Vo3aPz6S/ovGW5Xk3e+PiStuH3T+mdsYOz9E2G+vayGmN1RBDqiCDUEUGoI4JQRwShjghCHRGk1vuO7hVxG09+ir0mP2DyivcdF6pwXGVzbDJ5s8mvV+Exs9m59Cwzywaz2zn8YRFe4bZMd5u8DlgdEYQ6Igh1RBDqiCDUEUGoI4JQRwSp9b5jFer+i8mnmXy0ybeZMzXmjVKpedPgdjP5SPMcaMvbOu83W4RTm8zs20x+cJHOb1BhB3M3ZRJWRwShjghCHRGEOiIIdUQQ6ogg1BFBqrXveKXJe1VpfmGVydc36PzIezrvr87UeFC+8K6UxrU6b3H3R+pTMsrO6SKcaLYRy08m32Hu7WxV4Sjz/d0T5fU4hZjVEUGoI4JQRwShjghCHRGEOiIIdUSQau073mLyblWaX7lMx4fdjZBmk+7Qzyo1tx4+9JrOl/+q89Z3zMW4XHPbt+I9jla3/Tp/3Iy/r4K5q4XVEUGoI4JQRwShjghCHRGEOiJItTZ6+lY4fmcFY7uafI7J3aaTOQ91hnrp3wB9vGkZZnaubjSf+a7JK+S+qtsA+kqF5vu7k2/rgdURQagjglBHBKGOCEIdEYQ6Igh1RJBav1DP2VzBWPfkpdtfdG7TcYt6F546CaMU/7Tnyyb/oKjHZkvZUm5uG75Y7peDT5SbzPTOsypsrnCSOmB1RBDqiCDUEUGoI4JQRwShjghCHRGkXvuOPSoY22hyc9ZG2dDf5O5vb0Tb6JtygRz6QOki88XmOdDx5g9+fDnaNjzTnEO8pPwh8yb339dxqwjlW/aK/zXWA6sjglBHBKGOCEIdEYQ6Igh1RBDqiCDV2ncUu2illFLMeRhlgY5HPynChgrnXuNOFXE7bH+q8Dc5dGXZJfOvyysybyz6ge0RRdxlOav8IAdfVPbJfKZ5X+H+1t0idb+u7SavB1ZHBKGOCEIdEYQ6Igh1RBDqiCDUEUGqte/4iMn3mnyIjuU+mr7DsJQfTf7md+YHq00utxI/NYO1z01+t8lvVeHiMlEOntT1XJn3mfGtzEfKdL25lCdMXg+sjghCHRGEOiIIdUQQ6ogg1BFBav1g6/M1nFs/7WkPW62LZpMvVeGiruIR21JK5+l6kqHTVsh8uLof7UP3fLA5a7YuWB0RhDoiCHVEEOqIINQRQagjglBHBKnXC/WqYU29L+Af3LO0d5p8YIMIO08zo+/R8a7Vq2T+0VgzTzxWRwShjghCHRGEOiIIdUQQ6ogg1BFBwvYd5Svv3Hvwbjf51CpcSF+TP2ryMSbv/Zj5gbyFsbsee3yZzpsm6Nwd2ZGP1RFBqCOCUEcEoY4IQh0RhDoiCHVEkLB9R3nYhDuBorfJ55nh+qiNcu8BEU6+Vg++eLz5UHNosTmHuBT10r8t6/TYmS/prdc1pY+ZfaPJ07E6Igh1RBDqiCDUEUGoI4JQRwQJ2+ipRK9OOu9nTnGYbx73vPywSt0WSoU2mS2X2eocg9VP68Gtdq/rdFtNTrfvg1MadUQQ6ogg1BFBqCOCUEcEoY4IkrXv2EPt0i3crAc3DNT5pW52d0Nar/Yv6l8+UXejlVIal+t8nXnK9mgFn9mOwSZ/tTrT/+dYHRGEOiIIdUQQ6ogg1BFBqCOCUEcEqe2+4zUmdwdWDFIn355/hxk9yeTmuF1HbgHOnasHz1qg85Y9lX1ohdw7BU83rI4IQh0RhDoiCHVEEOqIINQRQagjgtR239Edb+FyrUnHX8zQ+VsmP26mn6PC5vYuqHbWmvyu//Qq6ofVEUGoI4JQRwShjghCHRGEOiIIdQQAAAAAAAAAAAAAAAAAAAAAAAAAAADwf/MXzCbbbwplbmRzdHJlYW0KZW5kb2JqCjIxIDAgb2JqCjE3NTEKZW5kb2JqCjIgMCBvYmoKPDwgL0NvdW50IDEgL0tpZHMgWyAxMCAwIFIgXSAvVHlwZSAvUGFnZXMgPj4KZW5kb2JqCjIyIDAgb2JqCjw8IC9DcmVhdGlvbkRhdGUgKEQ6MjAyMTAzMjIxMDU2NTItMDQnMDAnKQovQ3JlYXRvciAoTWF0cGxvdGxpYiB2My4zLjIsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcpCi9Qcm9kdWNlciAoTWF0cGxvdGxpYiBwZGYgYmFja2VuZCB2My4zLjIpID4+CmVuZG9iagp4cmVmCjAgMjMKMDAwMDAwMDAwMCA2NTUzNSBmIAowMDAwMDAwMDE2IDAwMDAwIG4gCjAwMDAwMDU5NjIgMDAwMDAgbiAKMDAwMDAwMzc1NiAwMDAwMCBuIAowMDAwMDAzNzg4IDAwMDAwIG4gCjAwMDAwMDM4ODcgMDAwMDAgbiAKMDAwMDAwMzkwOCAwMDAwMCBuIAowMDAwMDAzOTI5IDAwMDAwIG4gCjAwMDAwMDAwNjUgMDAwMDAgbiAKMDAwMDAwMDM5OCAwMDAwMCBuIAowMDAwMDAwMjA4IDAwMDAwIG4gCjAwMDAwMDEwMTQgMDAwMDAgbiAKMDAwMDAwMzk2MSAwMDAwMCBuIAowMDAwMDAyNjMxIDAwMDAwIG4gCjAwMDAwMDI0MzEgMDAwMDAgbiAKMDAwMDAwMjExMCAwMDAwMCBuIAowMDAwMDAzNjg0IDAwMDAwIG4gCjAwMDAwMDEwMzQgMDAwMDAgbiAKMDAwMDAwMTM1NCAwMDAwMCBuIAowMDAwMDAxNTA2IDAwMDAwIG4gCjAwMDAwMDE4MjcgMDAwMDAgbiAKMDAwMDAwNTk0MSAwMDAwMCBuIAowMDAwMDA2MDIyIDAwMDAwIG4gCnRyYWlsZXIKPDwgL0luZm8gMjIgMCBSIC9Sb290IDEgMCBSIC9TaXplIDIzID4+CnN0YXJ0eHJlZgo2MTc5CiUlRU9GCg==\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#grab 3 images\n",
    "x1, x2, x3 = mnist_data_train[0], mnist_data_train[1], mnist_data_train[2]\n",
    "#drop the labels\n",
    "x1, x2, x3 = x1[0], x2[0], x3[0]\n",
    "x_as_color = torch.stack([x1[0,:], x2[0,:], x3[0,:]], dim=0)\n",
    "imshow(x_as_color.permute(1,2,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.258941Z",
     "start_time": "2021-03-22T14:56:52.094057Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6a1fb72cd0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rand_order = torch.randperm(x_example.shape[1] * x_example.shape[2])\n",
    "x_shuffled = x_example.view(-1)[rand_order].view(x_example.shape)\n",
    "imshow(x_shuffled[0,:], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.486229Z",
     "start_time": "2021-03-22T14:56:52.260269Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f6a1f963b50>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy.signal import convolve\n",
    "img_indx = 58\n",
    "img = mnist_data_train[img_indx][0][0,:]\n",
    "plt.imshow(img, vmin=0, vmax=1, cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.653303Z",
     "start_time": "2021-03-22T14:56:52.487509Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "blur_filter = np.asarray([[1,1,1],\n",
    "                          [1,1,1],\n",
    "                          [1,1,1]\n",
    "                         ])/9.0\n",
    "\n",
    "blurry_img = convolve(img, blur_filter)\n",
    "plt.imshow(blurry_img, vmin=0, vmax=1, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.819553Z",
     "start_time": "2021-03-22T14:56:52.654427Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#We can find edges by focusing on the difference between a pixel, and its neighbors\n",
    "edge_filter = np.asarray([[-1,-1,-1],\n",
    "                          [-1, 8,-1],\n",
    "                          [-1,-1,-1]\n",
    "                         ])\n",
    "\n",
    "\n",
    "edge_img = convolve(img, edge_filter)\n",
    "plt.imshow(edge_img, vmin=0, vmax=1, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:52.986823Z",
     "start_time": "2021-03-22T14:56:52.822341Z"
    },
    "max_h": 0.3
   },
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#We could look for only horizontal edges\n",
    "h_edge_filter = np.asarray([[-1,-1,-1],\n",
    "                          [0, 0,0],\n",
    "                          [1, 1, 1]\n",
    "                         ])\n",
    "\n",
    "\n",
    "h_edge_img = convolve(img, h_edge_filter)\n",
    "plt.imshow(h_edge_img, vmin=0, vmax=1, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:53.024397Z",
     "start_time": "2021-03-22T14:56:52.988368Z"
    }
   },
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda\") if torch.cuda.is_available() else torch.device(\"cpu\")\n",
    "\n",
    "B = 32\n",
    "mnist_train_loader = DataLoader(mnist_data_train, batch_size=B, shuffle=True)\n",
    "mnist_test_loader = DataLoader(mnist_data_test, batch_size=B)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T14:56:53.032073Z",
     "start_time": "2021-03-22T14:56:53.025573Z"
    }
   },
   "outputs": [],
   "source": [
    "#How many values are in the input? We use this to help determine the size of subsequent layers\n",
    "D = 28*28 #28 * 28 images \n",
    "#How many channels are in the input?\n",
    "C = 1\n",
    "#How many classes are there?\n",
    "classes = 10\n",
    "#How many filters should we use\n",
    "filters = 16\n",
    "#how large should our filters be?\n",
    "K = 3\n",
    "#for comparison, lets define a linear model of similar complexity\n",
    "model_linear = nn.Sequential(\n",
    "  nn.Flatten(), #(B, C, W, H) -> (B, C*W*H) = (B,D)\n",
    "  nn.Linear(D, 256), \n",
    "  nn.Tanh(),\n",
    "  nn.Linear(256, classes),\n",
    ")\n",
    "\n",
    "#A simple convolutional network:\n",
    "model_cnn = nn.Sequential(\n",
    "  #Conv2d follows the pattern of:\n",
    "  #Conv2d(# of input channels, #filters/output-channels, #filter-size)\n",
    "  nn.Conv2d(C, filters, K, padding=K//2), #$x \\circledast G$\n",
    "  nn.Tanh(),#Activation functions work on any size tensor\n",
    "  nn.Flatten(), #Convert from (B, C, W, H) ->(B, D). This way we can use a Linear layer after\n",
    "  nn.Linear(filters*D, classes),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:02:46.846007Z",
     "start_time": "2021-03-22T14:56:53.033558Z"
    },
    "tags": [
     "remove_output"
    ]
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e6994c9db82b4721a70bea84463f0561",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Epoch'), FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ebcf1038023e43f2957a4d0a636fb685",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Epoch'), FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "loss_func = nn.CrossEntropyLoss()\n",
    "cnn_results = train_simple_network(model_cnn, loss_func, mnist_train_loader, test_loader=mnist_test_loader, score_funcs={'Accuracy': accuracy_score}, device=device, epochs=20)\n",
    "fc_results = train_simple_network(model_linear, loss_func, mnist_train_loader, test_loader=mnist_test_loader, score_funcs={'Accuracy': accuracy_score}, device=device, epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:02:47.104806Z",
     "start_time": "2021-03-22T15:02:46.847188Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='epoch', ylabel='test Accuracy'>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(x='epoch', y='test Accuracy', data=cnn_results, label='CNN')\n",
    "sns.lineplot(x='epoch', y='test Accuracy', data=fc_results, label='Fully Conected')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:02:47.425221Z",
     "start_time": "2021-03-22T15:02:47.110382Z"
    },
    "max_h": 0.4
   },
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img_indx = 0\n",
    "img, correct_class = mnist_data_train[img_indx]\n",
    "img = img[0,:]\n",
    "#move to the lower right, then upper left\n",
    "img_lr = np.roll(np.roll(img, 1, axis=1), 1, axis=0)\n",
    "img_ul = np.roll(np.roll(img, -1, axis=1), -1, axis=0)\n",
    "#plot the images\n",
    "f, axarr = plt.subplots(1,3)\n",
    "axarr[0].imshow(img, cmap='gray')\n",
    "axarr[1].imshow(img_lr, cmap='gray')\n",
    "axarr[2].imshow(img_ul, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:02:47.430933Z",
     "start_time": "2021-03-22T15:02:47.426453Z"
    }
   },
   "outputs": [],
   "source": [
    "#eval mode since we are not training\n",
    "model = model_cnn.cpu().eval()\n",
    "\n",
    "def pred(model, img):\n",
    "    with torch.no_grad():#Always turn off gradients when evaluating\n",
    "        w, h = img.shape#Whats the width/height of the image\n",
    "        if not isinstance(img, torch.Tensor):\n",
    "            img = torch.tensor(img)\n",
    "        x = img.reshape(1,-1,w,h)#reshape it as (B, C, W, H)\n",
    "        logits = model(x) #Get the logits\n",
    "        y_hat = F.softmax(logits, dim=1)#Turn into probabilities \n",
    "        return y_hat.numpy().flatten()#convert prediction to numpy array. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:02:47.449121Z",
     "start_time": "2021-03-22T15:02:47.432158Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Org Img Class 5 Prob:          0.78159285\n",
      "Lower Right Img Class 5 Prob:  0.44280732\n",
      "Uper Left Img Class 5 Prob:    0.31534675\n"
     ]
    }
   ],
   "source": [
    "img_pred = pred(model, img)\n",
    "img_lr_pred = pred(model, img_lr)\n",
    "img_ul_pred = pred(model, img_ul)\n",
    "\n",
    "print(\"Org Img Class {} Prob:         \".format(correct_class) , img_pred[correct_class])\n",
    "print(\"Lower Right Img Class {} Prob: \".format(correct_class) , img_lr_pred[correct_class])\n",
    "print(\"Uper Left Img Class {} Prob:   \".format(correct_class) , img_ul_pred[correct_class])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:09:01.336880Z",
     "start_time": "2021-03-22T15:02:47.468427Z"
    },
    "tags": [
     "remove_output"
    ]
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "50d7b59432544f31a33074911796c90d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Epoch'), FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model_cnn_pool = nn.Sequential(\n",
    "  nn.Conv2d(C, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(filters, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(filters, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.MaxPool2d(2),\n",
    "  nn.Conv2d(filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(2*filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(2*filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.MaxPool2d(2),\n",
    "\n",
    "  nn.Flatten(), \n",
    "  #Why did we reduce the number of units into the Linear layer by a factor of $4^2$? Because pooling a 2x2 grid down to one value means we go from 4 values, down to 1, and we did this two times. \n",
    "  nn.Linear(2*filters*D//(4**2), classes),\n",
    ")\n",
    "\n",
    "cnn_results_with_pool = train_simple_network(model_cnn_pool, loss_func, mnist_train_loader, test_loader=mnist_test_loader, score_funcs={'Accuracy': accuracy_score}, device=device, epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:09:01.353894Z",
     "start_time": "2021-03-22T15:09:01.338110Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Org Img Class 5 Prob:          0.7068047\n",
      "Lower Right Img Class 5 Prob:  0.71668524\n",
      "Uper Left Img Class 5 Prob:    0.7311974\n"
     ]
    }
   ],
   "source": [
    "model = model_cnn_pool.cpu().eval()\n",
    "img_pred = pred(model, img)\n",
    "img_lr_pred = pred(model, img_lr)\n",
    "img_ul_pred = pred(model, img_ul)\n",
    "\n",
    "print(\"Org Img Class {} Prob:         \".format(correct_class) , img_pred[correct_class])\n",
    "print(\"Lower Right Img Class {} Prob: \".format(correct_class) , img_lr_pred[correct_class])\n",
    "print(\"Uper Left Img Class {} Prob:   \".format(correct_class) , img_ul_pred[correct_class])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:09:01.598559Z",
     "start_time": "2021-03-22T15:09:01.355242Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='epoch', ylabel='test Accuracy'>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(x='epoch', y='test Accuracy', data=cnn_results, label='Simple CNN')\n",
    "sns.lineplot(x='epoch', y='test Accuracy', data=cnn_results_with_pool, label='CNN w/ Max Pooling')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:09:02.472082Z",
     "start_time": "2021-03-22T15:09:01.599867Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Several built-in transformations, given some agressive values to make their impact more obvious.  \n",
    "sample_transforms = {\n",
    "    \"Rotation\" : transforms.RandomAffine(degrees=45),\n",
    "    \"Translation\" : transforms.RandomAffine(degrees=0, translate=(0.1,0.1)),\n",
    "    \"Shear\": transforms.RandomAffine(degrees=0, shear=45),\n",
    "    \"RandomCrop\" : transforms.RandomCrop((20,20)),\n",
    "    \"Horizontal Flip\" : transforms.RandomHorizontalFlip(p=1.0),\n",
    "    \"Vertical Flip\": transforms.RandomVerticalFlip(p=1.0),\n",
    "    \"Perspective\": transforms.RandomPerspective(p=1.0),   \n",
    "    \"ColorJitter\" : transforms.ColorJitter(brightness=0.9, contrast=0.9)\n",
    "}\n",
    "#Convert the Tensor image back to a PIL image using a transform\n",
    "pil_img = transforms.ToPILImage()(img)\n",
    "#Plot a randomy application of each transform\n",
    "f, axarr = plt.subplots(2,4)\n",
    "for count, (name, t) in enumerate(sample_transforms.items()):\n",
    "    row = count % 4\n",
    "    col = count // 4\n",
    "    axarr[col,row].imshow(t(pil_img), cmap='gray')\n",
    "    axarr[col,row].set_title(name)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:09:02.499210Z",
     "start_time": "2021-03-22T15:09:02.473350Z"
    }
   },
   "outputs": [],
   "source": [
    "train_transform = transforms.Compose([\n",
    "    transforms.RandomAffine(degrees=5, translate=(0.05, 0.05), scale=(0.98, 1.02)),\n",
    "    transforms.ToTensor(),\n",
    "])\n",
    "\n",
    "test_transform = transforms.ToTensor()\n",
    "\n",
    "mnist_train_t = torchvision.datasets.MNIST(\"./data\", train=True, transform=train_transform)\n",
    "mnist_test_t = torchvision.datasets.MNIST(\"./data\", train=False, transform=test_transform)\n",
    "mnist_train_loader_t = DataLoader(mnist_train_t, shuffle=True,  batch_size=B, num_workers=5)\n",
    "mnist_test_loader_t = DataLoader(mnist_test_t, batch_size=B, num_workers=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:13:44.875053Z",
     "start_time": "2021-03-22T15:09:02.500926Z"
    },
    "tags": [
     "remove_output"
    ]
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5fab06dbe8264247838ea9f2098ca6ee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Epoch'), FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Training'), FloatProgress(value=0.0, max=1875.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Testing'), FloatProgress(value=0.0, max=313.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model_cnn_pool = nn.Sequential(\n",
    "  nn.Conv2d(C, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(filters, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(filters, filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.MaxPool2d(2),\n",
    "  nn.Conv2d(filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(2*filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.Conv2d(2*filters, 2*filters, 3, padding=3//2), \n",
    "  nn.Tanh(),\n",
    "  nn.MaxPool2d(2),\n",
    "  nn.Flatten(), \n",
    "  nn.Linear(2*filters*D//(4**2), classes),\n",
    ")\n",
    "\n",
    "cnn_results_with_pool_augmented = train_simple_network(model_cnn_pool, loss_func, mnist_train_loader_t, test_loader=mnist_test_loader_t, score_funcs={'Accuracy': accuracy_score}, device=device, epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-03-22T15:13:45.117991Z",
     "start_time": "2021-03-22T15:13:44.876560Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='epoch', ylabel='test Accuracy'>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/pdf": "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\n",
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(x='epoch', y='test Accuracy', data=cnn_results_with_pool, label='CNN w/ Max Pooling')\n",
    "sns.lineplot(x='epoch', y='test Accuracy', data=cnn_results_with_pool_augmented, label='CNN w/ Max Pooling + Augmentation')"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Tags",
  "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": "3.7.9"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": false,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": false,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "latex_metadata": {
   "title": "Convolutional Neural Networks"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}