Perceptron.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import torch.nn as nn\n",
    "from sklearn import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100,)\n"
     ]
    }
   ],
   "source": [
    "  n_pts = 100\n",
    "  centers = [[-0.5, 0.5], [0.5, -0.5]]\n",
    "  X, y = datasets.make_blobs(n_samples=n_pts, random_state=123, centers=centers, cluster_std=0.4)\n",
    "  \n",
    "  x_data = torch.Tensor(X)\n",
    "  y_data = torch.Tensor(y.reshape(100, 1))\n",
    "  print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def scatter_plot():\n",
    "    plt.scatter(X[y==0, 0], X[y==0, 1]) # Blue\n",
    "    plt.scatter(X[y==1, 0], X[y==1, 1]) # Orange"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "scatter_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model(nn.Module):\n",
    "    def __init__(self, input_size, output_size):\n",
    "      super().__init__() \n",
    "      self.linear = nn.Linear(input_size, output_size)\n",
    "    def forward(self, x):\n",
    "      pred = torch.sigmoid(self.linear(x))\n",
    "      return pred\n",
    "    def predict(self, x):\n",
    "      pred = self.forward(x)\n",
    "      if pred >= 0.5:\n",
    "        return 1\n",
    "      else:\n",
    "        return 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Parameter containing:\n",
      "tensor([[ 0.1622, -0.1683]], requires_grad=True), Parameter containing:\n",
      "tensor([0.1939], requires_grad=True)]\n"
     ]
    }
   ],
   "source": [
    "torch.manual_seed(2)\n",
    "model = Model(2, 1)\n",
    "print(list(model.parameters()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "[w, b] = model.parameters()\n",
    "w1, w2 = w.view(2)\n",
    "def get_params():\n",
    "  return (w1.item(), w2.item(), b[0].item())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_fit(title):\n",
    "    plt.title = title\n",
    "    w1, w2, b1 = get_params()\n",
    "    x1 = np.array([-2.0, 2.0])\n",
    "    x2 = -1*(w1*x1 + b1)/w2\n",
    "    scatter_plot()\n",
    "    plt.plot(x1, x2, 'r')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_fit('Initial Model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "criterion = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 0 loss: 0.6185115575790405\n",
      "epoch: 1 loss: 0.617464005947113\n",
      "epoch: 2 loss: 0.6164200305938721\n",
      "epoch: 3 loss: 0.6153795123100281\n",
      "epoch: 4 loss: 0.6143423914909363\n",
      "epoch: 5 loss: 0.6133086681365967\n",
      "epoch: 6 loss: 0.6122783422470093\n",
      "epoch: 7 loss: 0.6112514734268188\n",
      "epoch: 8 loss: 0.6102279424667358\n",
      "epoch: 9 loss: 0.6092078685760498\n",
      "epoch: 10 loss: 0.6081911325454712\n",
      "epoch: 11 loss: 0.6071776747703552\n",
      "epoch: 12 loss: 0.606167733669281\n",
      "epoch: 13 loss: 0.6051609516143799\n",
      "epoch: 14 loss: 0.604157567024231\n",
      "epoch: 15 loss: 0.6031573414802551\n",
      "epoch: 16 loss: 0.6021605730056763\n",
      "epoch: 17 loss: 0.6011670231819153\n",
      "epoch: 18 loss: 0.6001767516136169\n",
      "epoch: 19 loss: 0.599189817905426\n",
      "epoch: 20 loss: 0.5982059240341187\n",
      "epoch: 21 loss: 0.5972253680229187\n",
      "epoch: 22 loss: 0.5962480306625366\n",
      "epoch: 23 loss: 0.5952739119529724\n",
      "epoch: 24 loss: 0.5943029522895813\n",
      "epoch: 25 loss: 0.5933352112770081\n",
      "epoch: 26 loss: 0.5923706293106079\n",
      "epoch: 27 loss: 0.5914091467857361\n",
      "epoch: 28 loss: 0.5904508829116821\n",
      "epoch: 29 loss: 0.5894956588745117\n",
      "epoch: 30 loss: 0.5885436534881592\n",
      "epoch: 31 loss: 0.5875946283340454\n",
      "epoch: 32 loss: 0.5866488814353943\n",
      "epoch: 33 loss: 0.5857061147689819\n",
      "epoch: 34 loss: 0.5847664475440979\n",
      "epoch: 35 loss: 0.5838297605514526\n",
      "epoch: 36 loss: 0.5828962922096252\n",
      "epoch: 37 loss: 0.5819656848907471\n",
      "epoch: 38 loss: 0.5810381770133972\n",
      "epoch: 39 loss: 0.5801136493682861\n",
      "epoch: 40 loss: 0.5791922211647034\n",
      "epoch: 41 loss: 0.5782737135887146\n",
      "epoch: 42 loss: 0.5773581862449646\n",
      "epoch: 43 loss: 0.5764455795288086\n",
      "epoch: 44 loss: 0.5755359530448914\n",
      "epoch: 45 loss: 0.5746293067932129\n",
      "epoch: 46 loss: 0.5737256407737732\n",
      "epoch: 47 loss: 0.5728247761726379\n",
      "epoch: 48 loss: 0.5719268918037415\n",
      "epoch: 49 loss: 0.5710319876670837\n",
      "epoch: 50 loss: 0.5701398253440857\n",
      "epoch: 51 loss: 0.5692507028579712\n",
      "epoch: 52 loss: 0.5683643221855164\n",
      "epoch: 53 loss: 0.5674808621406555\n",
      "epoch: 54 loss: 0.5666002035140991\n",
      "epoch: 55 loss: 0.5657224059104919\n",
      "epoch: 56 loss: 0.564847469329834\n",
      "epoch: 57 loss: 0.5639752745628357\n",
      "epoch: 58 loss: 0.5631057620048523\n",
      "epoch: 59 loss: 0.5622392892837524\n",
      "epoch: 60 loss: 0.5613754391670227\n",
      "epoch: 61 loss: 0.5605143904685974\n",
      "epoch: 62 loss: 0.5596560835838318\n",
      "epoch: 63 loss: 0.5588005781173706\n",
      "epoch: 64 loss: 0.5579478144645691\n",
      "epoch: 65 loss: 0.5570976734161377\n",
      "epoch: 66 loss: 0.5562503337860107\n",
      "epoch: 67 loss: 0.5554057955741882\n",
      "epoch: 68 loss: 0.5545638203620911\n",
      "epoch: 69 loss: 0.5537245273590088\n",
      "epoch: 70 loss: 0.5528878569602966\n",
      "epoch: 71 loss: 0.5520539283752441\n",
      "epoch: 72 loss: 0.5512226819992065\n",
      "epoch: 73 loss: 0.5503940582275391\n",
      "epoch: 74 loss: 0.5495679974555969\n",
      "epoch: 75 loss: 0.5487446784973145\n",
      "epoch: 76 loss: 0.5479238629341125\n",
      "epoch: 77 loss: 0.5471056699752808\n",
      "epoch: 78 loss: 0.5462901592254639\n",
      "epoch: 79 loss: 0.5454770922660828\n",
      "epoch: 80 loss: 0.5446667075157166\n",
      "epoch: 81 loss: 0.5438588261604309\n",
      "epoch: 82 loss: 0.5430534482002258\n",
      "epoch: 83 loss: 0.5422507524490356\n",
      "epoch: 84 loss: 0.5414505004882812\n",
      "epoch: 85 loss: 0.5406527519226074\n",
      "epoch: 86 loss: 0.5398575663566589\n",
      "epoch: 87 loss: 0.539064884185791\n",
      "epoch: 88 loss: 0.5382745862007141\n",
      "epoch: 89 loss: 0.5374868512153625\n",
      "epoch: 90 loss: 0.5367016196250916\n",
      "epoch: 91 loss: 0.5359187126159668\n",
      "epoch: 92 loss: 0.5351383686065674\n",
      "epoch: 93 loss: 0.5343605279922485\n",
      "epoch: 94 loss: 0.5335850119590759\n",
      "epoch: 95 loss: 0.5328119397163391\n",
      "epoch: 96 loss: 0.5320413112640381\n",
      "epoch: 97 loss: 0.5312731266021729\n",
      "epoch: 98 loss: 0.5305072069168091\n",
      "epoch: 99 loss: 0.5297437906265259\n",
      "epoch: 100 loss: 0.5289827585220337\n",
      "epoch: 101 loss: 0.5282240509986877\n",
      "epoch: 102 loss: 0.5274677276611328\n",
      "epoch: 103 loss: 0.5267137289047241\n",
      "epoch: 104 loss: 0.5259620547294617\n",
      "epoch: 105 loss: 0.5252127051353455\n",
      "epoch: 106 loss: 0.5244657397270203\n",
      "epoch: 107 loss: 0.523720920085907\n",
      "epoch: 108 loss: 0.5229786038398743\n",
      "epoch: 109 loss: 0.5222384929656982\n",
      "epoch: 110 loss: 0.5215007662773132\n",
      "epoch: 111 loss: 0.5207651257514954\n",
      "epoch: 112 loss: 0.5200318694114685\n",
      "epoch: 113 loss: 0.5193009376525879\n",
      "epoch: 114 loss: 0.5185720920562744\n",
      "epoch: 115 loss: 0.517845630645752\n",
      "epoch: 116 loss: 0.5171212553977966\n",
      "epoch: 117 loss: 0.5163992047309875\n",
      "epoch: 118 loss: 0.5156794190406799\n",
      "epoch: 119 loss: 0.5149617791175842\n",
      "epoch: 120 loss: 0.5142463445663452\n",
      "epoch: 121 loss: 0.5135329961776733\n",
      "epoch: 122 loss: 0.5128219723701477\n",
      "epoch: 123 loss: 0.5121130347251892\n",
      "epoch: 124 loss: 0.5114063024520874\n",
      "epoch: 125 loss: 0.5107016563415527\n",
      "epoch: 126 loss: 0.5099992156028748\n",
      "epoch: 127 loss: 0.5092989206314087\n",
      "epoch: 128 loss: 0.5086007714271545\n",
      "epoch: 129 loss: 0.5079046487808228\n",
      "epoch: 130 loss: 0.5072107315063477\n",
      "epoch: 131 loss: 0.5065188407897949\n",
      "epoch: 132 loss: 0.5058290958404541\n",
      "epoch: 133 loss: 0.5051414966583252\n",
      "epoch: 134 loss: 0.5044559240341187\n",
      "epoch: 135 loss: 0.5037723779678345\n",
      "epoch: 136 loss: 0.5030909180641174\n",
      "epoch: 137 loss: 0.5024115443229675\n",
      "epoch: 138 loss: 0.50173419713974\n",
      "epoch: 139 loss: 0.5010589361190796\n",
      "epoch: 140 loss: 0.5003856420516968\n",
      "epoch: 141 loss: 0.4997144639492035\n",
      "epoch: 142 loss: 0.4990452229976654\n",
      "epoch: 143 loss: 0.4983779489994049\n",
      "epoch: 144 loss: 0.4977127015590668\n",
      "epoch: 145 loss: 0.497049480676651\n",
      "epoch: 146 loss: 0.4963882565498352\n",
      "epoch: 147 loss: 0.4957289397716522\n",
      "epoch: 148 loss: 0.4950715899467468\n",
      "epoch: 149 loss: 0.4944162666797638\n",
      "epoch: 150 loss: 0.4937628209590912\n",
      "epoch: 151 loss: 0.4931114614009857\n",
      "epoch: 152 loss: 0.4924619793891907\n",
      "epoch: 153 loss: 0.4918143153190613\n",
      "epoch: 154 loss: 0.4911685883998871\n",
      "epoch: 155 loss: 0.49052494764328003\n",
      "epoch: 156 loss: 0.489883154630661\n",
      "epoch: 157 loss: 0.48924314975738525\n",
      "epoch: 158 loss: 0.4886051118373871\n",
      "epoch: 159 loss: 0.48796898126602173\n",
      "epoch: 160 loss: 0.48733463883399963\n",
      "epoch: 161 loss: 0.48670223355293274\n",
      "epoch: 162 loss: 0.48607170581817627\n",
      "epoch: 163 loss: 0.48544302582740784\n",
      "epoch: 164 loss: 0.48481619358062744\n",
      "epoch: 165 loss: 0.4841911792755127\n",
      "epoch: 166 loss: 0.4835680425167084\n",
      "epoch: 167 loss: 0.4829466640949249\n",
      "epoch: 168 loss: 0.4823271632194519\n",
      "epoch: 169 loss: 0.48170945048332214\n",
      "epoch: 170 loss: 0.48109352588653564\n",
      "epoch: 171 loss: 0.4804794192314148\n",
      "epoch: 172 loss: 0.479867160320282\n",
      "epoch: 173 loss: 0.47925662994384766\n",
      "epoch: 174 loss: 0.47864779829978943\n",
      "epoch: 175 loss: 0.478040874004364\n",
      "epoch: 176 loss: 0.4774357080459595\n",
      "epoch: 177 loss: 0.47683221101760864\n",
      "epoch: 178 loss: 0.47623053193092346\n",
      "epoch: 179 loss: 0.475630521774292\n",
      "epoch: 180 loss: 0.47503235936164856\n",
      "epoch: 181 loss: 0.47443583607673645\n",
      "epoch: 182 loss: 0.4738411009311676\n",
      "epoch: 183 loss: 0.47324806451797485\n",
      "epoch: 184 loss: 0.4726566970348358\n",
      "epoch: 185 loss: 0.47206705808639526\n",
      "epoch: 186 loss: 0.4714791178703308\n",
      "epoch: 187 loss: 0.47089287638664246\n",
      "epoch: 188 loss: 0.4703083336353302\n",
      "epoch: 189 loss: 0.46972543001174927\n",
      "epoch: 190 loss: 0.46914416551589966\n",
      "epoch: 191 loss: 0.46856462955474854\n",
      "epoch: 192 loss: 0.4679867625236511\n",
      "epoch: 193 loss: 0.46741050481796265\n",
      "epoch: 194 loss: 0.46683594584465027\n",
      "epoch: 195 loss: 0.4662629961967468\n",
      "epoch: 196 loss: 0.4656916856765747\n",
      "epoch: 197 loss: 0.4651219844818115\n",
      "epoch: 198 loss: 0.46455392241477966\n",
      "epoch: 199 loss: 0.46398743987083435\n",
      "epoch: 200 loss: 0.46342262625694275\n",
      "epoch: 201 loss: 0.4628593921661377\n",
      "epoch: 202 loss: 0.4622976779937744\n",
      "epoch: 203 loss: 0.46173763275146484\n",
      "epoch: 204 loss: 0.4611791968345642\n",
      "epoch: 205 loss: 0.46062228083610535\n",
      "epoch: 206 loss: 0.4600669741630554\n",
      "epoch: 207 loss: 0.4595131278038025\n",
      "epoch: 208 loss: 0.45896098017692566\n",
      "epoch: 209 loss: 0.4584103524684906\n",
      "epoch: 210 loss: 0.45786112546920776\n",
      "epoch: 211 loss: 0.4573136568069458\n",
      "epoch: 212 loss: 0.45676758885383606\n",
      "epoch: 213 loss: 0.4562230706214905\n",
      "epoch: 214 loss: 0.45568013191223145\n",
      "epoch: 215 loss: 0.4551387131214142\n",
      "epoch: 216 loss: 0.45459872484207153\n",
      "epoch: 217 loss: 0.45406031608581543\n",
      "epoch: 218 loss: 0.45352333784103394\n",
      "epoch: 219 loss: 0.4529879093170166\n",
      "epoch: 220 loss: 0.45245397090911865\n",
      "epoch: 221 loss: 0.4519214928150177\n",
      "epoch: 222 loss: 0.45139050483703613\n",
      "epoch: 223 loss: 0.45086100697517395\n",
      "epoch: 224 loss: 0.4503329396247864\n",
      "epoch: 225 loss: 0.4498063325881958\n",
      "epoch: 226 loss: 0.44928115606307983\n",
      "epoch: 227 loss: 0.44875743985176086\n",
      "epoch: 228 loss: 0.4482351541519165\n",
      "epoch: 229 loss: 0.4477144479751587\n",
      "epoch: 230 loss: 0.4471951425075531\n",
      "epoch: 231 loss: 0.4466770887374878\n",
      "epoch: 232 loss: 0.4461606740951538\n",
      "epoch: 233 loss: 0.4456455707550049\n",
      "epoch: 234 loss: 0.4451318681240082\n",
      "epoch: 235 loss: 0.4446195662021637\n",
      "epoch: 236 loss: 0.4441086947917938\n",
      "epoch: 237 loss: 0.4435991942882538\n",
      "epoch: 238 loss: 0.44309109449386597\n",
      "epoch: 239 loss: 0.44258439540863037\n",
      "epoch: 240 loss: 0.4420790374279022\n",
      "epoch: 241 loss: 0.44157519936561584\n",
      "epoch: 242 loss: 0.44107258319854736\n",
      "epoch: 243 loss: 0.4405713379383087\n",
      "epoch: 244 loss: 0.4400714635848999\n",
      "epoch: 245 loss: 0.43957293033599854\n",
      "epoch: 246 loss: 0.43907585740089417\n",
      "epoch: 247 loss: 0.4385799765586853\n",
      "epoch: 248 loss: 0.43808555603027344\n",
      "epoch: 249 loss: 0.43759241700172424\n",
      "epoch: 250 loss: 0.4371006488800049\n",
      "epoch: 251 loss: 0.4366101324558258\n",
      "epoch: 252 loss: 0.4361209571361542\n",
      "epoch: 253 loss: 0.4356330931186676\n",
      "epoch: 254 loss: 0.43514660000801086\n",
      "epoch: 255 loss: 0.4346614181995392\n",
      "epoch: 256 loss: 0.434177428483963\n",
      "epoch: 257 loss: 0.43369483947753906\n",
      "epoch: 258 loss: 0.433213472366333\n",
      "epoch: 259 loss: 0.4327333867549896\n",
      "epoch: 260 loss: 0.4322546720504761\n",
      "epoch: 261 loss: 0.43177714943885803\n",
      "epoch: 262 loss: 0.43130093812942505\n",
      "epoch: 263 loss: 0.4308259189128876\n",
      "epoch: 264 loss: 0.4303523004055023\n",
      "epoch: 265 loss: 0.4298798441886902\n",
      "epoch: 266 loss: 0.42940858006477356\n",
      "epoch: 267 loss: 0.428938627243042\n",
      "epoch: 268 loss: 0.42847001552581787\n",
      "epoch: 269 loss: 0.4280024766921997\n",
      "epoch: 270 loss: 0.4275362491607666\n",
      "epoch: 271 loss: 0.42707130312919617\n",
      "epoch: 272 loss: 0.42660748958587646\n",
      "epoch: 273 loss: 0.42614495754241943\n",
      "epoch: 274 loss: 0.42568355798721313\n",
      "epoch: 275 loss: 0.4252234995365143\n",
      "epoch: 276 loss: 0.42476460337638855\n",
      "epoch: 277 loss: 0.42430686950683594\n",
      "epoch: 278 loss: 0.42385032773017883\n",
      "epoch: 279 loss: 0.4233950078487396\n",
      "epoch: 280 loss: 0.42294085025787354\n",
      "epoch: 281 loss: 0.4224879741668701\n",
      "epoch: 282 loss: 0.42203617095947266\n",
      "epoch: 283 loss: 0.4215856194496155\n",
      "epoch: 284 loss: 0.4211360812187195\n",
      "epoch: 285 loss: 0.4206879138946533\n",
      "epoch: 286 loss: 0.4202407896518707\n",
      "epoch: 287 loss: 0.41979485750198364\n",
      "epoch: 288 loss: 0.41935011744499207\n",
      "epoch: 289 loss: 0.41890645027160645\n",
      "epoch: 290 loss: 0.41846397519111633\n",
      "epoch: 291 loss: 0.41802266240119934\n",
      "epoch: 292 loss: 0.4175824820995331\n",
      "epoch: 293 loss: 0.41714340448379517\n",
      "epoch: 294 loss: 0.41670548915863037\n",
      "epoch: 295 loss: 0.4162687063217163\n",
      "epoch: 296 loss: 0.4158330261707306\n",
      "epoch: 297 loss: 0.41539841890335083\n",
      "epoch: 298 loss: 0.4149649739265442\n",
      "epoch: 299 loss: 0.4145326316356659\n",
      "epoch: 300 loss: 0.41410136222839355\n",
      "epoch: 301 loss: 0.41367119550704956\n",
      "epoch: 302 loss: 0.4132421910762787\n",
      "epoch: 303 loss: 0.41281428933143616\n",
      "epoch: 304 loss: 0.4123874604701996\n",
      "epoch: 305 loss: 0.4119616448879242\n",
      "epoch: 306 loss: 0.4115369915962219\n",
      "epoch: 307 loss: 0.41111335158348083\n",
      "epoch: 308 loss: 0.4106908142566681\n",
      "epoch: 309 loss: 0.4102693498134613\n",
      "epoch: 310 loss: 0.40984898805618286\n",
      "epoch: 311 loss: 0.4094296395778656\n",
      "epoch: 312 loss: 0.4090113341808319\n",
      "epoch: 313 loss: 0.40859413146972656\n",
      "epoch: 314 loss: 0.4081779420375824\n",
      "epoch: 315 loss: 0.40776288509368896\n",
      "epoch: 316 loss: 0.40734875202178955\n",
      "epoch: 317 loss: 0.4069356918334961\n",
      "epoch: 318 loss: 0.4065236747264862\n",
      "epoch: 319 loss: 0.40611276030540466\n",
      "epoch: 320 loss: 0.40570276975631714\n",
      "epoch: 321 loss: 0.40529391169548035\n",
      "epoch: 322 loss: 0.4048859477043152\n",
      "epoch: 323 loss: 0.40447911620140076\n",
      "epoch: 324 loss: 0.40407320857048035\n",
      "epoch: 325 loss: 0.4036683142185211\n",
      "epoch: 326 loss: 0.40326449275016785\n",
      "epoch: 327 loss: 0.40286168456077576\n",
      "epoch: 328 loss: 0.4024598300457001\n",
      "epoch: 329 loss: 0.40205898880958557\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 330 loss: 0.40165916085243225\n",
      "epoch: 331 loss: 0.40126025676727295\n",
      "epoch: 332 loss: 0.40086233615875244\n",
      "epoch: 333 loss: 0.4004655182361603\n",
      "epoch: 334 loss: 0.40006953477859497\n",
      "epoch: 335 loss: 0.39967459440231323\n",
      "epoch: 336 loss: 0.3992806375026703\n",
      "epoch: 337 loss: 0.39888760447502136\n",
      "epoch: 338 loss: 0.3984955847263336\n",
      "epoch: 339 loss: 0.3981045186519623\n",
      "epoch: 340 loss: 0.39771437644958496\n",
      "epoch: 341 loss: 0.3973252475261688\n",
      "epoch: 342 loss: 0.3969370126724243\n",
      "epoch: 343 loss: 0.3965497314929962\n",
      "epoch: 344 loss: 0.3961634337902069\n",
      "epoch: 345 loss: 0.3957780599594116\n",
      "epoch: 346 loss: 0.39539361000061035\n",
      "epoch: 347 loss: 0.3950101137161255\n",
      "epoch: 348 loss: 0.39462754130363464\n",
      "epoch: 349 loss: 0.3942459225654602\n",
      "epoch: 350 loss: 0.3938651978969574\n",
      "epoch: 351 loss: 0.393485426902771\n",
      "epoch: 352 loss: 0.3931065499782562\n",
      "epoch: 353 loss: 0.3927285373210907\n",
      "epoch: 354 loss: 0.39235153794288635\n",
      "epoch: 355 loss: 0.39197537302970886\n",
      "epoch: 356 loss: 0.391600102186203\n",
      "epoch: 357 loss: 0.3912258446216583\n",
      "epoch: 358 loss: 0.39085233211517334\n",
      "epoch: 359 loss: 0.39047980308532715\n",
      "epoch: 360 loss: 0.3901081383228302\n",
      "epoch: 361 loss: 0.38973739743232727\n",
      "epoch: 362 loss: 0.38936755061149597\n",
      "epoch: 363 loss: 0.38899853825569153\n",
      "epoch: 364 loss: 0.3886304497718811\n",
      "epoch: 365 loss: 0.3882632553577423\n",
      "epoch: 366 loss: 0.38789692521095276\n",
      "epoch: 367 loss: 0.38753142952919006\n",
      "epoch: 368 loss: 0.387166827917099\n",
      "epoch: 369 loss: 0.3868030905723572\n",
      "epoch: 370 loss: 0.3864401876926422\n",
      "epoch: 371 loss: 0.38607823848724365\n",
      "epoch: 372 loss: 0.3857170045375824\n",
      "epoch: 373 loss: 0.38535675406455994\n",
      "epoch: 374 loss: 0.38499727845191956\n",
      "epoch: 375 loss: 0.3846386969089508\n",
      "epoch: 376 loss: 0.3842809796333313\n",
      "epoch: 377 loss: 0.38392403721809387\n",
      "epoch: 378 loss: 0.3835679590702057\n",
      "epoch: 379 loss: 0.3832126557826996\n",
      "epoch: 380 loss: 0.3828582465648651\n",
      "epoch: 381 loss: 0.38250476121902466\n",
      "epoch: 382 loss: 0.3821519911289215\n",
      "epoch: 383 loss: 0.3818000853061676\n",
      "epoch: 384 loss: 0.38144898414611816\n",
      "epoch: 385 loss: 0.3810987174510956\n",
      "epoch: 386 loss: 0.3807491660118103\n",
      "epoch: 387 loss: 0.3804006278514862\n",
      "epoch: 388 loss: 0.38005271553993225\n",
      "epoch: 389 loss: 0.3797057271003723\n",
      "epoch: 390 loss: 0.3793594241142273\n",
      "epoch: 391 loss: 0.37901410460472107\n",
      "epoch: 392 loss: 0.3786694407463074\n",
      "epoch: 393 loss: 0.3783256411552429\n",
      "epoch: 394 loss: 0.37798258662223816\n",
      "epoch: 395 loss: 0.37764033675193787\n",
      "epoch: 396 loss: 0.37729892134666443\n",
      "epoch: 397 loss: 0.37695828080177307\n",
      "epoch: 398 loss: 0.3766184151172638\n",
      "epoch: 399 loss: 0.3762793242931366\n",
      "epoch: 400 loss: 0.37594109773635864\n",
      "epoch: 401 loss: 0.37560340762138367\n",
      "epoch: 402 loss: 0.37526676058769226\n",
      "epoch: 403 loss: 0.37493079900741577\n",
      "epoch: 404 loss: 0.3745955228805542\n",
      "epoch: 405 loss: 0.3742610812187195\n",
      "epoch: 406 loss: 0.37392738461494446\n",
      "epoch: 407 loss: 0.3735944628715515\n",
      "epoch: 408 loss: 0.37326228618621826\n",
      "epoch: 409 loss: 0.3729309141635895\n",
      "epoch: 410 loss: 0.3726001977920532\n",
      "epoch: 411 loss: 0.372270405292511\n",
      "epoch: 412 loss: 0.3719411790370941\n",
      "epoch: 413 loss: 0.3716128170490265\n",
      "epoch: 414 loss: 0.3712851405143738\n",
      "epoch: 415 loss: 0.37095823884010315\n",
      "epoch: 416 loss: 0.37063202261924744\n",
      "epoch: 417 loss: 0.3703065812587738\n",
      "epoch: 418 loss: 0.36998191475868225\n",
      "epoch: 419 loss: 0.3696579337120056\n",
      "epoch: 420 loss: 0.3693346083164215\n",
      "epoch: 421 loss: 0.36901211738586426\n",
      "epoch: 422 loss: 0.3686903715133667\n",
      "epoch: 423 loss: 0.36836928129196167\n",
      "epoch: 424 loss: 0.36804890632629395\n",
      "epoch: 425 loss: 0.3677293062210083\n",
      "epoch: 426 loss: 0.3674103617668152\n",
      "epoch: 427 loss: 0.3670921325683594\n",
      "epoch: 428 loss: 0.3667745590209961\n",
      "epoch: 429 loss: 0.36645781993865967\n",
      "epoch: 430 loss: 0.36614176630973816\n",
      "epoch: 431 loss: 0.3658263385295868\n",
      "epoch: 432 loss: 0.36551162600517273\n",
      "epoch: 433 loss: 0.36519771814346313\n",
      "epoch: 434 loss: 0.3648844063282013\n",
      "epoch: 435 loss: 0.36457180976867676\n",
      "epoch: 436 loss: 0.36425989866256714\n",
      "epoch: 437 loss: 0.36394864320755005\n",
      "epoch: 438 loss: 0.36363816261291504\n",
      "epoch: 439 loss: 0.36332833766937256\n",
      "epoch: 440 loss: 0.3630191385746002\n",
      "epoch: 441 loss: 0.3627106547355652\n",
      "epoch: 442 loss: 0.36240288615226746\n",
      "epoch: 443 loss: 0.3620957136154175\n",
      "epoch: 444 loss: 0.3617892861366272\n",
      "epoch: 445 loss: 0.36148345470428467\n",
      "epoch: 446 loss: 0.3611783981323242\n",
      "epoch: 447 loss: 0.3608739376068115\n",
      "epoch: 448 loss: 0.36057019233703613\n",
      "epoch: 449 loss: 0.3602670431137085\n",
      "epoch: 450 loss: 0.35996466875076294\n",
      "epoch: 451 loss: 0.35966286063194275\n",
      "epoch: 452 loss: 0.3593616783618927\n",
      "epoch: 453 loss: 0.35906124114990234\n",
      "epoch: 454 loss: 0.35876137018203735\n",
      "epoch: 455 loss: 0.35846224427223206\n",
      "epoch: 456 loss: 0.3581637144088745\n",
      "epoch: 457 loss: 0.3578658401966095\n",
      "epoch: 458 loss: 0.3575685918331146\n",
      "epoch: 459 loss: 0.35727202892303467\n",
      "epoch: 460 loss: 0.35697606205940247\n",
      "epoch: 461 loss: 0.3566807210445404\n",
      "epoch: 462 loss: 0.3563860356807709\n",
      "epoch: 463 loss: 0.35609200596809387\n",
      "epoch: 464 loss: 0.35579854249954224\n",
      "epoch: 465 loss: 0.35550570487976074\n",
      "epoch: 466 loss: 0.3552136719226837\n",
      "epoch: 467 loss: 0.3549221158027649\n",
      "epoch: 468 loss: 0.3546311557292938\n",
      "epoch: 469 loss: 0.35434091091156006\n",
      "epoch: 470 loss: 0.35405123233795166\n",
      "epoch: 471 loss: 0.3537622392177582\n",
      "epoch: 472 loss: 0.3534737825393677\n",
      "epoch: 473 loss: 0.3531860113143921\n",
      "epoch: 474 loss: 0.3528987169265747\n",
      "epoch: 475 loss: 0.35261213779449463\n",
      "epoch: 476 loss: 0.3523261547088623\n",
      "epoch: 477 loss: 0.3520408272743225\n",
      "epoch: 478 loss: 0.3517560064792633\n",
      "epoch: 479 loss: 0.351471871137619\n",
      "epoch: 480 loss: 0.3511883020401001\n",
      "epoch: 481 loss: 0.35090529918670654\n",
      "epoch: 482 loss: 0.3506229519844055\n",
      "epoch: 483 loss: 0.35034120082855225\n",
      "epoch: 484 loss: 0.35006001591682434\n",
      "epoch: 485 loss: 0.3497793972492218\n",
      "epoch: 486 loss: 0.3494994342327118\n",
      "epoch: 487 loss: 0.3492199182510376\n",
      "epoch: 488 loss: 0.3489411175251007\n",
      "epoch: 489 loss: 0.3486629128456116\n",
      "epoch: 490 loss: 0.3483852744102478\n",
      "epoch: 491 loss: 0.3481081426143646\n",
      "epoch: 492 loss: 0.34783169627189636\n",
      "epoch: 493 loss: 0.34755581617355347\n",
      "epoch: 494 loss: 0.347280353307724\n",
      "epoch: 495 loss: 0.34700557589530945\n",
      "epoch: 496 loss: 0.34673136472702026\n",
      "epoch: 497 loss: 0.3464577794075012\n",
      "epoch: 498 loss: 0.34618470072746277\n",
      "epoch: 499 loss: 0.34591221809387207\n",
      "epoch: 500 loss: 0.34564027190208435\n",
      "epoch: 501 loss: 0.345368891954422\n",
      "epoch: 502 loss: 0.3450981378555298\n",
      "epoch: 503 loss: 0.344827800989151\n",
      "epoch: 504 loss: 0.34455814957618713\n",
      "epoch: 505 loss: 0.34428900480270386\n",
      "epoch: 506 loss: 0.34402039647102356\n",
      "epoch: 507 loss: 0.34375235438346863\n",
      "epoch: 508 loss: 0.34348487854003906\n",
      "epoch: 509 loss: 0.3432179391384125\n",
      "epoch: 510 loss: 0.34295153617858887\n",
      "epoch: 511 loss: 0.342685729265213\n",
      "epoch: 512 loss: 0.34242039918899536\n",
      "epoch: 513 loss: 0.3421556353569031\n",
      "epoch: 514 loss: 0.34189146757125854\n",
      "epoch: 515 loss: 0.3416277766227722\n",
      "epoch: 516 loss: 0.34136462211608887\n",
      "epoch: 517 loss: 0.3411020338535309\n",
      "epoch: 518 loss: 0.3408399224281311\n",
      "epoch: 519 loss: 0.34057843685150146\n",
      "epoch: 520 loss: 0.3403174579143524\n",
      "epoch: 521 loss: 0.3400569558143616\n",
      "epoch: 522 loss: 0.3397969901561737\n",
      "epoch: 523 loss: 0.3395376205444336\n",
      "epoch: 524 loss: 0.3392786681652069\n",
      "epoch: 525 loss: 0.33902034163475037\n",
      "epoch: 526 loss: 0.3387625217437744\n",
      "epoch: 527 loss: 0.3385050892829895\n",
      "epoch: 528 loss: 0.3382483720779419\n",
      "epoch: 529 loss: 0.3379921019077301\n",
      "epoch: 530 loss: 0.3377362787723541\n",
      "epoch: 531 loss: 0.33748093247413635\n",
      "epoch: 532 loss: 0.3372262120246887\n",
      "epoch: 533 loss: 0.3369719982147217\n",
      "epoch: 534 loss: 0.33671826124191284\n",
      "epoch: 535 loss: 0.3364650011062622\n",
      "epoch: 536 loss: 0.33621230721473694\n",
      "epoch: 537 loss: 0.3359600901603699\n",
      "epoch: 538 loss: 0.3357083797454834\n",
      "epoch: 539 loss: 0.3354571461677551\n",
      "epoch: 540 loss: 0.33520641922950745\n",
      "epoch: 541 loss: 0.33495616912841797\n",
      "epoch: 542 loss: 0.33470645546913147\n",
      "epoch: 543 loss: 0.33445724844932556\n",
      "epoch: 544 loss: 0.33420857787132263\n",
      "epoch: 545 loss: 0.3339603841304779\n",
      "epoch: 546 loss: 0.3337125778198242\n",
      "epoch: 547 loss: 0.3334653079509735\n",
      "epoch: 548 loss: 0.3332185745239258\n",
      "epoch: 549 loss: 0.3329722583293915\n",
      "epoch: 550 loss: 0.33272647857666016\n",
      "epoch: 551 loss: 0.33248111605644226\n",
      "epoch: 552 loss: 0.33223631978034973\n",
      "epoch: 553 loss: 0.3319920003414154\n",
      "epoch: 554 loss: 0.3317481279373169\n",
      "epoch: 555 loss: 0.3315047323703766\n",
      "epoch: 556 loss: 0.3312617838382721\n",
      "epoch: 557 loss: 0.33101943135261536\n",
      "epoch: 558 loss: 0.3307774066925049\n",
      "epoch: 559 loss: 0.3305359184741974\n",
      "epoch: 560 loss: 0.3302949070930481\n",
      "epoch: 561 loss: 0.33005428314208984\n",
      "epoch: 562 loss: 0.32981425523757935\n",
      "epoch: 563 loss: 0.32957467436790466\n",
      "epoch: 564 loss: 0.3293355107307434\n",
      "epoch: 565 loss: 0.3290967643260956\n",
      "epoch: 566 loss: 0.32885855436325073\n",
      "epoch: 567 loss: 0.3286207914352417\n",
      "epoch: 568 loss: 0.32838353514671326\n",
      "epoch: 569 loss: 0.32814663648605347\n",
      "epoch: 570 loss: 0.32791027426719666\n",
      "epoch: 571 loss: 0.32767435908317566\n",
      "epoch: 572 loss: 0.3274388611316681\n",
      "epoch: 573 loss: 0.3272038400173187\n",
      "epoch: 574 loss: 0.3269692659378052\n",
      "epoch: 575 loss: 0.32673513889312744\n",
      "epoch: 576 loss: 0.3265014886856079\n",
      "epoch: 577 loss: 0.3262682259082794\n",
      "epoch: 578 loss: 0.3260354995727539\n",
      "epoch: 579 loss: 0.32580316066741943\n",
      "epoch: 580 loss: 0.3255712389945984\n",
      "epoch: 581 loss: 0.32533982396125793\n",
      "epoch: 582 loss: 0.3251087963581085\n",
      "epoch: 583 loss: 0.3248782753944397\n",
      "epoch: 584 loss: 0.3246481418609619\n",
      "epoch: 585 loss: 0.32441845536231995\n",
      "epoch: 586 loss: 0.3241891860961914\n",
      "epoch: 587 loss: 0.32396042346954346\n",
      "epoch: 588 loss: 0.32373207807540894\n",
      "epoch: 589 loss: 0.32350409030914307\n",
      "epoch: 590 loss: 0.3232765197753906\n",
      "epoch: 591 loss: 0.32304951548576355\n",
      "epoch: 592 loss: 0.3228228688240051\n",
      "epoch: 593 loss: 0.32259663939476013\n",
      "epoch: 594 loss: 0.32237082719802856\n",
      "epoch: 595 loss: 0.3221454918384552\n",
      "epoch: 596 loss: 0.3219205439090729\n",
      "epoch: 597 loss: 0.321696013212204\n",
      "epoch: 598 loss: 0.3214718997478485\n",
      "epoch: 599 loss: 0.32124823331832886\n",
      "epoch: 600 loss: 0.321025013923645\n",
      "epoch: 601 loss: 0.3208021819591522\n",
      "epoch: 602 loss: 0.3205797076225281\n",
      "epoch: 603 loss: 0.32035771012306213\n",
      "epoch: 604 loss: 0.32013610005378723\n",
      "epoch: 605 loss: 0.31991493701934814\n",
      "epoch: 606 loss: 0.3196941614151001\n",
      "epoch: 607 loss: 0.3194738030433655\n",
      "epoch: 608 loss: 0.3192538917064667\n",
      "epoch: 609 loss: 0.3190343379974365\n",
      "epoch: 610 loss: 0.3188151717185974\n",
      "epoch: 611 loss: 0.3185964822769165\n",
      "epoch: 612 loss: 0.31837818026542664\n",
      "epoch: 613 loss: 0.3181602358818054\n",
      "epoch: 614 loss: 0.31794273853302\n",
      "epoch: 615 loss: 0.31772568821907043\n",
      "epoch: 616 loss: 0.3175089359283447\n",
      "epoch: 617 loss: 0.3172926604747772\n",
      "epoch: 618 loss: 0.31707674264907837\n",
      "epoch: 619 loss: 0.31686124205589294\n",
      "epoch: 620 loss: 0.3166460692882538\n",
      "epoch: 621 loss: 0.3164314031600952\n",
      "epoch: 622 loss: 0.3162170648574829\n",
      "epoch: 623 loss: 0.31600314378738403\n",
      "epoch: 624 loss: 0.3157896101474762\n",
      "epoch: 625 loss: 0.3155764639377594\n",
      "epoch: 626 loss: 0.31536370515823364\n",
      "epoch: 627 loss: 0.3151513338088989\n",
      "epoch: 628 loss: 0.31493937969207764\n",
      "epoch: 629 loss: 0.3147277534008026\n",
      "epoch: 630 loss: 0.314516544342041\n",
      "epoch: 631 loss: 0.31430569291114807\n",
      "epoch: 632 loss: 0.31409525871276855\n",
      "epoch: 633 loss: 0.3138852119445801\n",
      "epoch: 634 loss: 0.31367552280426025\n",
      "epoch: 635 loss: 0.31346625089645386\n",
      "epoch: 636 loss: 0.3132573068141937\n",
      "epoch: 637 loss: 0.313048779964447\n",
      "epoch: 638 loss: 0.31284064054489136\n",
      "epoch: 639 loss: 0.31263285875320435\n",
      "epoch: 640 loss: 0.312425434589386\n",
      "epoch: 641 loss: 0.31221842765808105\n",
      "epoch: 642 loss: 0.3120117485523224\n",
      "epoch: 643 loss: 0.3118054270744324\n",
      "epoch: 644 loss: 0.3115995526313782\n",
      "epoch: 645 loss: 0.31139397621154785\n",
      "epoch: 646 loss: 0.31118878722190857\n",
      "epoch: 647 loss: 0.3109840154647827\n",
      "epoch: 648 loss: 0.3107795715332031\n",
      "epoch: 649 loss: 0.3105754852294922\n",
      "epoch: 650 loss: 0.3103717863559723\n",
      "epoch: 651 loss: 0.31016844511032104\n",
      "epoch: 652 loss: 0.30996549129486084\n",
      "epoch: 653 loss: 0.3097628355026245\n",
      "epoch: 654 loss: 0.3095605969429016\n",
      "epoch: 655 loss: 0.309358686208725\n",
      "epoch: 656 loss: 0.309157133102417\n",
      "epoch: 657 loss: 0.30895593762397766\n",
      "epoch: 658 loss: 0.308755099773407\n",
      "epoch: 659 loss: 0.30855464935302734\n",
      "epoch: 660 loss: 0.30835452675819397\n",
      "epoch: 661 loss: 0.30815476179122925\n",
      "epoch: 662 loss: 0.30795538425445557\n",
      "epoch: 663 loss: 0.30775633454322815\n",
      "epoch: 664 loss: 0.3075576424598694\n",
      "epoch: 665 loss: 0.3073592782020569\n",
      "epoch: 666 loss: 0.3071613013744354\n",
      "epoch: 667 loss: 0.30696359276771545\n",
      "epoch: 668 loss: 0.3067663311958313\n",
      "epoch: 669 loss: 0.30656933784484863\n",
      "epoch: 670 loss: 0.3063727617263794\n",
      "epoch: 671 loss: 0.30617645382881165\n",
      "epoch: 672 loss: 0.3059805631637573\n",
      "epoch: 673 loss: 0.3057849705219269\n",
      "epoch: 674 loss: 0.3055897355079651\n",
      "epoch: 675 loss: 0.30539488792419434\n",
      "epoch: 676 loss: 0.3052003085613251\n",
      "epoch: 677 loss: 0.30500608682632446\n",
      "epoch: 678 loss: 0.30481216311454773\n",
      "epoch: 679 loss: 0.30461862683296204\n",
      "epoch: 680 loss: 0.3044254779815674\n",
      "epoch: 681 loss: 0.30423256754875183\n",
      "epoch: 682 loss: 0.3040400445461273\n",
      "epoch: 683 loss: 0.3038478195667267\n",
      "epoch: 684 loss: 0.3036559820175171\n",
      "epoch: 685 loss: 0.3034643828868866\n",
      "epoch: 686 loss: 0.3032732307910919\n",
      "epoch: 687 loss: 0.30308234691619873\n",
      "epoch: 688 loss: 0.3028918206691742\n",
      "epoch: 689 loss: 0.30270159244537354\n",
      "epoch: 690 loss: 0.30251169204711914\n",
      "epoch: 691 loss: 0.302322119474411\n",
      "epoch: 692 loss: 0.3021329343318939\n",
      "epoch: 693 loss: 0.30194395780563354\n",
      "epoch: 694 loss: 0.3017553985118866\n",
      "epoch: 695 loss: 0.3015671670436859\n",
      "epoch: 696 loss: 0.30137917399406433\n",
      "epoch: 697 loss: 0.3011915683746338\n",
      "epoch: 698 loss: 0.3010042905807495\n",
      "epoch: 699 loss: 0.3008173108100891\n",
      "epoch: 700 loss: 0.300630658864975\n",
      "epoch: 701 loss: 0.3004443049430847\n",
      "epoch: 702 loss: 0.3002582788467407\n",
      "epoch: 703 loss: 0.3000725507736206\n",
      "epoch: 704 loss: 0.29988721013069153\n",
      "epoch: 705 loss: 0.29970213770866394\n",
      "epoch: 706 loss: 0.2995173931121826\n",
      "epoch: 707 loss: 0.29933294653892517\n",
      "epoch: 708 loss: 0.2991487681865692\n",
      "epoch: 709 loss: 0.2989649474620819\n",
      "epoch: 710 loss: 0.29878148436546326\n",
      "epoch: 711 loss: 0.2985982596874237\n",
      "epoch: 712 loss: 0.2984153628349304\n",
      "epoch: 713 loss: 0.2982327938079834\n",
      "epoch: 714 loss: 0.29805052280426025\n",
      "epoch: 715 loss: 0.2978685796260834\n",
      "epoch: 716 loss: 0.297686904668808\n",
      "epoch: 717 loss: 0.29750555753707886\n",
      "epoch: 718 loss: 0.2973244786262512\n",
      "epoch: 719 loss: 0.29714375734329224\n",
      "epoch: 720 loss: 0.29696330428123474\n",
      "epoch: 721 loss: 0.2967832088470459\n",
      "epoch: 722 loss: 0.29660335183143616\n",
      "epoch: 723 loss: 0.2964237928390503\n",
      "epoch: 724 loss: 0.2962445914745331\n",
      "epoch: 725 loss: 0.29606565833091736\n",
      "epoch: 726 loss: 0.2958870232105255\n",
      "epoch: 727 loss: 0.29570868611335754\n",
      "epoch: 728 loss: 0.29553061723709106\n",
      "epoch: 729 loss: 0.29535290598869324\n",
      "epoch: 730 loss: 0.2951754331588745\n",
      "epoch: 731 loss: 0.29499828815460205\n",
      "epoch: 732 loss: 0.2948213815689087\n",
      "epoch: 733 loss: 0.294644832611084\n",
      "epoch: 734 loss: 0.29446855187416077\n",
      "epoch: 735 loss: 0.2942925691604614\n",
      "epoch: 736 loss: 0.29411688446998596\n",
      "epoch: 737 loss: 0.2939414978027344\n",
      "epoch: 738 loss: 0.2937663793563843\n",
      "epoch: 739 loss: 0.29359158873558044\n",
      "epoch: 740 loss: 0.2934170365333557\n",
      "epoch: 741 loss: 0.29324278235435486\n",
      "epoch: 742 loss: 0.2930687665939331\n",
      "epoch: 743 loss: 0.29289510846138\n",
      "epoch: 744 loss: 0.2927217185497284\n",
      "epoch: 745 loss: 0.29254862666130066\n",
      "epoch: 746 loss: 0.2923758029937744\n",
      "epoch: 747 loss: 0.29220321774482727\n",
      "epoch: 748 loss: 0.2920309603214264\n",
      "epoch: 749 loss: 0.2918590009212494\n",
      "epoch: 750 loss: 0.2916873097419739\n",
      "epoch: 751 loss: 0.29151591658592224\n",
      "epoch: 752 loss: 0.2913447916507721\n",
      "epoch: 753 loss: 0.29117390513420105\n",
      "epoch: 754 loss: 0.2910033166408539\n",
      "epoch: 755 loss: 0.2908330261707306\n",
      "epoch: 756 loss: 0.2906629741191864\n",
      "epoch: 757 loss: 0.2904932498931885\n",
      "epoch: 758 loss: 0.29032376408576965\n",
      "epoch: 759 loss: 0.2901545763015747\n",
      "epoch: 760 loss: 0.28998565673828125\n",
      "epoch: 761 loss: 0.2898169755935669\n",
      "epoch: 762 loss: 0.28964856266975403\n",
      "epoch: 763 loss: 0.2894804775714874\n",
      "epoch: 764 loss: 0.2893126904964447\n",
      "epoch: 765 loss: 0.2891450822353363\n",
      "epoch: 766 loss: 0.28897780179977417\n",
      "epoch: 767 loss: 0.28881075978279114\n",
      "epoch: 768 loss: 0.288644015789032\n",
      "epoch: 769 loss: 0.28847751021385193\n",
      "epoch: 770 loss: 0.28831133246421814\n",
      "epoch: 771 loss: 0.2881453335285187\n",
      "epoch: 772 loss: 0.2879796624183655\n",
      "epoch: 773 loss: 0.2878142297267914\n",
      "epoch: 774 loss: 0.28764909505844116\n",
      "epoch: 775 loss: 0.28748419880867004\n",
      "epoch: 776 loss: 0.287319540977478\n",
      "epoch: 777 loss: 0.2871551811695099\n",
      "epoch: 778 loss: 0.28699102997779846\n",
      "epoch: 779 loss: 0.2868272066116333\n",
      "epoch: 780 loss: 0.28666362166404724\n",
      "epoch: 781 loss: 0.2865002751350403\n",
      "epoch: 782 loss: 0.2863372266292572\n",
      "epoch: 783 loss: 0.2861744165420532\n",
      "epoch: 784 loss: 0.28601187467575073\n",
      "epoch: 785 loss: 0.28584960103034973\n",
      "epoch: 786 loss: 0.28568753600120544\n",
      "epoch: 787 loss: 0.28552576899528503\n",
      "epoch: 788 loss: 0.28536418080329895\n",
      "epoch: 789 loss: 0.2852029502391815\n",
      "epoch: 790 loss: 0.2850419282913208\n",
      "epoch: 791 loss: 0.28488120436668396\n",
      "epoch: 792 loss: 0.28472065925598145\n",
      "epoch: 793 loss: 0.2845603823661804\n",
      "epoch: 794 loss: 0.2844003736972809\n",
      "epoch: 795 loss: 0.28424063324928284\n",
      "epoch: 796 loss: 0.2840811610221863\n",
      "epoch: 797 loss: 0.28392189741134644\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 798 loss: 0.2837629020214081\n",
      "epoch: 799 loss: 0.28360411524772644\n",
      "epoch: 800 loss: 0.2834455966949463\n",
      "epoch: 801 loss: 0.28328737616539\n",
      "epoch: 802 loss: 0.28312933444976807\n",
      "epoch: 803 loss: 0.2829716205596924\n",
      "epoch: 804 loss: 0.282814085483551\n",
      "epoch: 805 loss: 0.28265678882598877\n",
      "epoch: 806 loss: 0.2824997305870056\n",
      "epoch: 807 loss: 0.2823430001735687\n",
      "epoch: 808 loss: 0.2821864187717438\n",
      "epoch: 809 loss: 0.2820301353931427\n",
      "epoch: 810 loss: 0.2818741202354431\n",
      "epoch: 811 loss: 0.28171828389167786\n",
      "epoch: 812 loss: 0.2815626859664917\n",
      "epoch: 813 loss: 0.2814073860645294\n",
      "epoch: 814 loss: 0.28125226497650146\n",
      "epoch: 815 loss: 0.2810974419116974\n",
      "epoch: 816 loss: 0.28094282746315\n",
      "epoch: 817 loss: 0.280788391828537\n",
      "epoch: 818 loss: 0.2806342840194702\n",
      "epoch: 819 loss: 0.28048038482666016\n",
      "epoch: 820 loss: 0.2803267240524292\n",
      "epoch: 821 loss: 0.28017327189445496\n",
      "epoch: 822 loss: 0.2800200879573822\n",
      "epoch: 823 loss: 0.27986711263656616\n",
      "epoch: 824 loss: 0.2797144055366516\n",
      "epoch: 825 loss: 0.2795619070529938\n",
      "epoch: 826 loss: 0.2794096767902374\n",
      "epoch: 827 loss: 0.2792576551437378\n",
      "epoch: 828 loss: 0.2791058421134949\n",
      "epoch: 829 loss: 0.27895426750183105\n",
      "epoch: 830 loss: 0.2788029909133911\n",
      "epoch: 831 loss: 0.2786518335342407\n",
      "epoch: 832 loss: 0.2785009443759918\n",
      "epoch: 833 loss: 0.2783503532409668\n",
      "epoch: 834 loss: 0.2781999111175537\n",
      "epoch: 835 loss: 0.2780497372150421\n",
      "epoch: 836 loss: 0.27789977192878723\n",
      "epoch: 837 loss: 0.27775007486343384\n",
      "epoch: 838 loss: 0.27760055661201477\n",
      "epoch: 839 loss: 0.2774512767791748\n",
      "epoch: 840 loss: 0.27730223536491394\n",
      "epoch: 841 loss: 0.2771534025669098\n",
      "epoch: 842 loss: 0.27700480818748474\n",
      "epoch: 843 loss: 0.2768564224243164\n",
      "epoch: 844 loss: 0.2767082452774048\n",
      "epoch: 845 loss: 0.27656033635139465\n",
      "epoch: 846 loss: 0.27641263604164124\n",
      "epoch: 847 loss: 0.2762651741504669\n",
      "epoch: 848 loss: 0.27611789107322693\n",
      "epoch: 849 loss: 0.2759708762168884\n",
      "epoch: 850 loss: 0.27582404017448425\n",
      "epoch: 851 loss: 0.2756773829460144\n",
      "epoch: 852 loss: 0.2755310535430908\n",
      "epoch: 853 loss: 0.2753848731517792\n",
      "epoch: 854 loss: 0.27523890137672424\n",
      "epoch: 855 loss: 0.2750932276248932\n",
      "epoch: 856 loss: 0.2749477028846741\n",
      "epoch: 857 loss: 0.27480238676071167\n",
      "epoch: 858 loss: 0.27465736865997314\n",
      "epoch: 859 loss: 0.27451246976852417\n",
      "epoch: 860 loss: 0.2743678390979767\n",
      "epoch: 861 loss: 0.2742234170436859\n",
      "epoch: 862 loss: 0.27407920360565186\n",
      "epoch: 863 loss: 0.2739351689815521\n",
      "epoch: 864 loss: 0.2737914025783539\n",
      "epoch: 865 loss: 0.27364781498908997\n",
      "epoch: 866 loss: 0.27350449562072754\n",
      "epoch: 867 loss: 0.27336129546165466\n",
      "epoch: 868 loss: 0.27321839332580566\n",
      "epoch: 869 loss: 0.273075670003891\n",
      "epoch: 870 loss: 0.27293315529823303\n",
      "epoch: 871 loss: 0.2727908194065094\n",
      "epoch: 872 loss: 0.27264872193336487\n",
      "epoch: 873 loss: 0.27250683307647705\n",
      "epoch: 874 loss: 0.27236515283584595\n",
      "epoch: 875 loss: 0.27222371101379395\n",
      "epoch: 876 loss: 0.2720824182033539\n",
      "epoch: 877 loss: 0.2719413638114929\n",
      "epoch: 878 loss: 0.27180051803588867\n",
      "epoch: 879 loss: 0.27165988087654114\n",
      "epoch: 880 loss: 0.2715194821357727\n",
      "epoch: 881 loss: 0.2713792026042938\n",
      "epoch: 882 loss: 0.27123919129371643\n",
      "epoch: 883 loss: 0.27109935879707336\n",
      "epoch: 884 loss: 0.2709597647190094\n",
      "epoch: 885 loss: 0.270820289850235\n",
      "epoch: 886 loss: 0.27068108320236206\n",
      "epoch: 887 loss: 0.27054205536842346\n",
      "epoch: 888 loss: 0.27040329575538635\n",
      "epoch: 889 loss: 0.2702646553516388\n",
      "epoch: 890 loss: 0.27012625336647034\n",
      "epoch: 891 loss: 0.2699880301952362\n",
      "epoch: 892 loss: 0.26985007524490356\n",
      "epoch: 893 loss: 0.2697122395038605\n",
      "epoch: 894 loss: 0.2695746123790741\n",
      "epoch: 895 loss: 0.26943719387054443\n",
      "epoch: 896 loss: 0.2692999839782715\n",
      "epoch: 897 loss: 0.26916301250457764\n",
      "epoch: 898 loss: 0.2690261900424957\n",
      "epoch: 899 loss: 0.26888954639434814\n",
      "epoch: 900 loss: 0.2687531113624573\n",
      "epoch: 901 loss: 0.2686168849468231\n",
      "epoch: 902 loss: 0.2684808671474457\n",
      "epoch: 903 loss: 0.2683449983596802\n",
      "epoch: 904 loss: 0.2682093679904938\n",
      "epoch: 905 loss: 0.2680739462375641\n",
      "epoch: 906 loss: 0.26793867349624634\n",
      "epoch: 907 loss: 0.2678036689758301\n",
      "epoch: 908 loss: 0.267668753862381\n",
      "epoch: 909 loss: 0.26753413677215576\n",
      "epoch: 910 loss: 0.2673996090888977\n",
      "epoch: 911 loss: 0.26726534962654114\n",
      "epoch: 912 loss: 0.2671312689781189\n",
      "epoch: 913 loss: 0.266997367143631\n",
      "epoch: 914 loss: 0.2668636441230774\n",
      "epoch: 915 loss: 0.2667301297187805\n",
      "epoch: 916 loss: 0.2665967643260956\n",
      "epoch: 917 loss: 0.26646360754966736\n",
      "epoch: 918 loss: 0.26633062958717346\n",
      "epoch: 919 loss: 0.26619791984558105\n",
      "epoch: 920 loss: 0.2660652697086334\n",
      "epoch: 921 loss: 0.26593291759490967\n",
      "epoch: 922 loss: 0.26580074429512024\n",
      "epoch: 923 loss: 0.26566869020462036\n",
      "epoch: 924 loss: 0.2655368745326996\n",
      "epoch: 925 loss: 0.26540520787239075\n",
      "epoch: 926 loss: 0.2652737498283386\n",
      "epoch: 927 loss: 0.2651424705982208\n",
      "epoch: 928 loss: 0.26501137018203735\n",
      "epoch: 929 loss: 0.264880508184433\n",
      "epoch: 930 loss: 0.26474976539611816\n",
      "epoch: 931 loss: 0.26461920142173767\n",
      "epoch: 932 loss: 0.2644888460636139\n",
      "epoch: 933 loss: 0.26435863971710205\n",
      "epoch: 934 loss: 0.2642287313938141\n",
      "epoch: 935 loss: 0.2640989124774933\n",
      "epoch: 936 loss: 0.2639692723751068\n",
      "epoch: 937 loss: 0.26383981108665466\n",
      "epoch: 938 loss: 0.26371055841445923\n",
      "epoch: 939 loss: 0.26358145475387573\n",
      "epoch: 940 loss: 0.26345255970954895\n",
      "epoch: 941 loss: 0.2633238434791565\n",
      "epoch: 942 loss: 0.263195276260376\n",
      "epoch: 943 loss: 0.2630668878555298\n",
      "epoch: 944 loss: 0.2629387080669403\n",
      "epoch: 945 loss: 0.26281073689460754\n",
      "epoch: 946 loss: 0.26268285512924194\n",
      "epoch: 947 loss: 0.26255524158477783\n",
      "epoch: 948 loss: 0.2624277174472809\n",
      "epoch: 949 loss: 0.2623004615306854\n",
      "epoch: 950 loss: 0.26217329502105713\n",
      "epoch: 951 loss: 0.26204636693000793\n",
      "epoch: 952 loss: 0.2619195878505707\n",
      "epoch: 953 loss: 0.26179295778274536\n",
      "epoch: 954 loss: 0.26166659593582153\n",
      "epoch: 955 loss: 0.2615402936935425\n",
      "epoch: 956 loss: 0.26141422986984253\n",
      "epoch: 957 loss: 0.2612883448600769\n",
      "epoch: 958 loss: 0.2611626386642456\n",
      "epoch: 959 loss: 0.2610369920730591\n",
      "epoch: 960 loss: 0.26091164350509644\n",
      "epoch: 961 loss: 0.26078641414642334\n",
      "epoch: 962 loss: 0.26066139340400696\n",
      "epoch: 963 loss: 0.2605365514755249\n",
      "epoch: 964 loss: 0.2604118287563324\n",
      "epoch: 965 loss: 0.2602872848510742\n",
      "epoch: 966 loss: 0.26016291975975037\n",
      "epoch: 967 loss: 0.26003873348236084\n",
      "epoch: 968 loss: 0.25991469621658325\n",
      "epoch: 969 loss: 0.2597908079624176\n",
      "epoch: 970 loss: 0.25966721773147583\n",
      "epoch: 971 loss: 0.25954365730285645\n",
      "epoch: 972 loss: 0.2594203054904938\n",
      "epoch: 973 loss: 0.25929713249206543\n",
      "epoch: 974 loss: 0.2591741383075714\n",
      "epoch: 975 loss: 0.25905129313468933\n",
      "epoch: 976 loss: 0.2589286267757416\n",
      "epoch: 977 loss: 0.25880610942840576\n",
      "epoch: 978 loss: 0.2586837410926819\n",
      "epoch: 979 loss: 0.25856152176856995\n",
      "epoch: 980 loss: 0.2584395110607147\n",
      "epoch: 981 loss: 0.2583177089691162\n",
      "epoch: 982 loss: 0.25819605588912964\n",
      "epoch: 983 loss: 0.2580745220184326\n",
      "epoch: 984 loss: 0.25795307755470276\n",
      "epoch: 985 loss: 0.2578319311141968\n",
      "epoch: 986 loss: 0.25771090388298035\n",
      "epoch: 987 loss: 0.25759002566337585\n",
      "epoch: 988 loss: 0.2574693262577057\n",
      "epoch: 989 loss: 0.25734877586364746\n",
      "epoch: 990 loss: 0.25722837448120117\n",
      "epoch: 991 loss: 0.2571081817150116\n",
      "epoch: 992 loss: 0.25698813796043396\n",
      "epoch: 993 loss: 0.25686824321746826\n",
      "epoch: 994 loss: 0.2567484974861145\n",
      "epoch: 995 loss: 0.2566288709640503\n",
      "epoch: 996 loss: 0.2565094828605652\n",
      "epoch: 997 loss: 0.25639021396636963\n",
      "epoch: 998 loss: 0.256271094083786\n",
      "epoch: 999 loss: 0.2561521828174591\n"
     ]
    }
   ],
   "source": [
    "epochs = 1000\n",
    "losses = []\n",
    "\n",
    "for i in range(epochs):\n",
    "  y_pred = model.forward(x_data)\n",
    "  loss = criterion(y_pred, y_data)\n",
    "  print(\"epoch:\", i, \"loss:\", loss.item())\n",
    "  losses.append(loss.item())\n",
    "  optimizer.zero_grad()\n",
    "  loss.backward()\n",
    "  optimizer.step()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(epochs), losses)\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('epochs')\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_fit(\"Trained Model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Red point positive probability = 0.9424158334732056\n",
      "Black point positive probability = 0.05055497586727142\n",
      "Red point belongs in class 1\n",
      "Black point belongs in class = 0\n"
     ]
    }
   ],
   "source": [
    "point1 = torch.tensor([1.0, -1.0])\n",
    "point2 = torch.tensor([-1.0, 1.0])\n",
    "plt.plot(point1.numpy()[0], point1.numpy()[1], 'go')\n",
    "plt.plot(point2.numpy()[0], point2.numpy()[1], 'ko')\n",
    "plot_fit(\"Trained Model\")\n",
    "\n",
    "print(\"Red point positive probability = {}\".format(model.forward(point1).item())) \n",
    "print(\"Black point positive probability = {}\".format(model.forward(point2).item())) \n",
    "print(\"Red point belongs in class {}\".format(model.predict(point1))) \n",
    "print(\"Black point belongs in class = {}\".format(model.predict(point2))) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}