Artificial Neural Network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Basis of Artificial Neural Network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### input neuron=1, hidden layer 1 =6, hidden layer 2 = 10, hidden layer 3 =6, output neuron =1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn  # torch module for ANN\n",
    "import torch.optim as optim  # torch module for optimizer algorithms\n",
    "import torch.nn.init as init  # torch module for initialize values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_data = 1000\n",
    "num_epoch = 10000\n",
    "\n",
    "noise = init.normal_(torch.FloatTensor(num_data, 1), std = 1)\n",
    "x = init.uniform_(torch.Tensor(num_data, 1), -15, 15)\n",
    "y = (x**2 + 3)\n",
    "y_noise = y + noise\n",
    "\n",
    "# build model, loss_func, optimizer\n",
    "model = nn.Sequential(\n",
    "        nn.Linear(1, 6),\n",
    "        nn.ReLU(),\n",
    "        nn.Linear(6, 10),\n",
    "        nn.ReLU(),\n",
    "        nn.Linear(10, 6),\n",
    "        nn.ReLU(),\n",
    "        nn.Linear(6, 1),\n",
    ")\n",
    "\n",
    "loss_func = nn.L1Loss()\n",
    "optimizer = optim.SGD(model.parameters(), lr = 0.0002)\n",
    "\n",
    "# training\n",
    "loss_array = []\n",
    "for i in range(num_epoch):\n",
    "    optimizer.zero_grad()\n",
    "    output = model(x)\n",
    "    loss = loss_func(output, y)\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    \n",
    "    loss_array.append(loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAD4CAYAAAAaT9YAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAeuElEQVR4nO3de3CcV53m8e9PavVFUreuLVmWfFPimFwgji0SwiUT8CSEwAYyBTNhl0yGAULtDAzD7O5smNndYmtqalkKlkttbXZMBjYMd0IypMIMScYQyFCJEyUOjuNL5Lsty1JLsqybdT/7R7+SZVuWWlK3+u3u51OlertPvy39juM8Oj79vueYcw4REclNRdkuQERElk4hLiKSwxTiIiI5TCEuIpLDFOIiIjkssJI/rLa21q1fv34lf6SISM576aWXup1z8bleW9EQX79+Pa2trSv5I0VEcp6ZHbvca5pOERHJYQpxEZEcphAXEclhCnERkRymEBcRyWEKcRGRHKYQFxHJYSt6nfhS/WJ/Jwe7BrlpQw3Xro4RKNbvHhERyJEQf+ZAgm8/l7zWvSxYzNb11WxZW8l1qyu4rrGC+lgIM8tylSIiK89WclOIlpYWt9Q7Nrv6R9h5pJcXjvSy80gPbV2DTJdeWx7k2tUVXNcYmwn2pqqIgl1E8oKZveSca5nztVwJ8YsNjU6w/3Q/e9r72dN+lj2n+mnrHGBiKtmf8lCA5ngZV8TLucI7NsfLWV9bSihQnJYaRERWwnwhnhPTKXMpCwXYuq6areuqZ9pGxid5vXOAPe39vN45wKHEIDsP9/DYrvaZc4oM1lSXzoR7c7x85nF1WVCjdxHJKTkb4nMJlxTzpqZK3tRUeUH78NgEhxNDHEoMcmj62DXIbw52MzoxNXNeZWkJV8TLub6pkhs3VPPm9VXUlIdWuBciIqnL2emUdJiacrT3nbsg3Ns6B9h98uxMuG+qj7Lt6jp+95p6NjdVUlSkkbqIrKy8nBPPpNGJSV49eZYXjvby7OvdvHC0l8kpx+qKMB/c2sSHWtawpro022WKSIFQiC/T2eFxfnGgk8d2neLZtgTOwbY31PFn2zZy/ZrKbJcnInlOIZ5Gp/rO8cMXT/Dwc0fpGx7n1k1x/st7r+bKumi2SxORPLWsEDezTcAPZzU1A/8NqAQ+ASS89r9yzv3TfN8rH0J82sDION9+7hjbf32Y4bEJ/v3vXMGfvutKXb4oImmXtpG4mRUD7cBNwEeBQefcl1J9fz6F+LTuwVH+9mf7eGxXO9c3VfB/PrKVxspItssSkTwyX4gvdhGSbcAh59xl93srNLXlIb7yB5v5u3u3cjgxxPu+/iwvHevNdlkiUiAWG+L3AN+f9fxTZrbbzL5pZlVzvcHM7jezVjNrTSQSc52SF9597Soe//TbqSwN8pGHXuBf27qzXZKIFICUQ9zMgsBdwI+9pgeBK4DNQAfw5bne55zb7pxrcc61xOPx5VXrcxtqy/jRJ29mXU0pH//2i+w6fibbJYlInlvMSPw9wMvOuU4A51ync27SOTcFfAO4MRMF5pp4NMR3Pn4TddEwH3+4leM9w9kuSUTy2GJC/MPMmkoxs4ZZr90N7ElXUbmutjzE//vomxmfnOLT33+ZsVm39ouIpFNKIW5mpcBtwKOzmr9oZq+a2W7gncBnM1BfzmqOl/PFD76J3548y5efPpDtckQkT6W0AJZzbhiouajt3oxUlEfuuK6BD9+4hm/8+jB3Xb+aa1dXZLskEckz2ucswx6442qqSoP813/cw0reHSsihUEhnmEVpSX85zvewMvH+3hqb2e2yxGRPKMQXwG/t6WRDbVlfPVf2pia0mhcRNJHIb4CAsVFfPpdV7Kvo58d+7uyXY6I5BGF+Aq56/rVrIqF+fZzR7NdiojkEYX4CgkUF/Fvb1rLs23dHOkeynY5IpInFOIr6J43ryFQZPzgxePZLkVE8oRCfAXVxcK8fWMtT/y2Q5cbikhaKMRX2PvetJr2vnPsOtGX7VJEJA8oxFfY7dfWEywu4onfdmS7FBHJAwrxFRYLl/C2K2v45QFdaigiy6cQz4JbropzpHtIy9SKyLIpxLPglquSm2P8qi1/dzoSkZWhEM+C5toymqoi/Pp1hbiILI9CPAvMjJuba2g92qtLDUVkWRTiWbJ1XRVnhsd196aILItCPEu2rqsC4KVj2kxZRJZOIZ4lV8TLiYUDvHxcIS4iS7dgiJvZJjN7ZdZXv5n9uZlVm9nTZtbmHatWouB8UVRk3LC2ipeP9WW7FBHJYQuGuHPugHNus3NuM7AVGAYeAx4AdjjnNgI7vOeyCNc1xjiYGGRkfDLbpYhIjlrsdMo24JBz7hjwfuBhr/1h4ANprKsgXN0QY3LKcbBrMNuliEiOWmyI3wN833tc75zrAPCOdXO9wczuN7NWM2tNJHRd9GxXN8QA2NfRn+VKRCRXpRziZhYE7gJ+vJgf4Jzb7pxrcc61xOPxxdaX19bXlBEuKWJfx0C2SxGRHLWYkfh7gJedc9NbtneaWQOAd9SKTotUXGRsWhXTSFxElmwxIf5hzk+lADwO3Oc9vg/4abqKKiSb6ss5mNCcuIgsTUohbmalwG3Ao7OavwDcZmZt3mtfSH95+W9DbTmJgVEGRsazXYqI5KBAKic554aBmovaekherSLLsKG2DICj3cO8sakiy9WISK7RHZtZ1hxPhvjhbk2piMjiKcSzbG11KWZoISwRWRKFeJaFS4pprIxwVCEuIkugEPeBDbVlGomLyJIoxH2gqSpCe9+5bJchIjlIIe4DjZURugfHtBCWiCyaQtwHVldGADil0biILJJC3AfOh/hIlisRkVyjEPeBRo3ERWSJFOI+sKoijBmcVIiLyCIpxH2gpLiI+mhYI3ERWTSFuE80VIbpOKsQF5HFUYj7RF00RFf/aLbLEJEcoxD3ibpomMSgQlxEFkch7hPxaIi+4XFGJ3TDj4ikTiHuE3XREADdg2NZrkREcolC3CfiXoh39euGHxFJnULcJ+qiYQASA5oXF5HUpbrHZqWZPWJm+81sn5ndbGafN7N2M3vF+7oz08Xms5mRuEJcRBYhpT02ga8BP3fOfdDMgkAp8G7gK865L2WsugJSUx7ETCNxEVmcBUPczGLALcAfATjnxoAxM8tsZQWmpLiI6tKgRuIisiipTKc0AwngW2a2y8weMrMy77VPmdluM/ummVXN9WYzu9/MWs2sNZFIpKvuvBSPhjQSF5FFSSXEA8AW4EHn3A3AEPAA8CBwBbAZ6AC+PNebnXPbnXMtzrmWeDyelqLzVTLEdXWKiKQulRA/CZx0zu30nj8CbHHOdTrnJp1zU8A3gBszVWShqC0P0TOk68RFJHULhrhz7jRwwsw2eU3bgL1m1jDrtLuBPRmor6BUlQY5oxAXkUVI9eqUTwPf9a5MOQx8FPi6mW0GHHAU+GQmCiwk1WUlDI1NMjI+SbikONvliEgOSCnEnXOvAC0XNd+b9moKXHVZ8lrxvuFxVlUoxEVkYbpj00eqy0oA6NWUioikSCHuI1WlQQDODCvERSQ1CnEfqS5LhriuUBGRVCnEfaTKC3FdoSIiqVKI+0hlRHPiIrI4CnEfCRQXUREp0Zy4iKRMIe4zNWVBjcRFJGUKcZ+pKgtqJC4iKVOI+0xVaZDeofFslyEiOUIh7jPVZSW6OkVEUqYQ95kqb07cOZftUkQkByjEfaYiUsLY5BSjE1PZLkVEcoBC3GcqvGvFz57TvLiILEwh7jOxsEJcRFKnEPcZjcRFZDEU4j4zE+LDCnERWZhC3GemQ7x/RCEuIgtTiPuMplNEZDFSCnEzqzSzR8xsv5ntM7ObzazazJ42szbvWJXpYgtBNJzcMU8hLiKpSHUk/jXg5865NwDXA/uAB4AdzrmNwA7vuSxToLiI8lBAIS4iKVkwxM0sBtwC/D2Ac27MOdcHvB942DvtYeADmSmx8FREShTiIpKSVEbizUAC+JaZ7TKzh8ysDKh3znUAeMe6ud5sZvebWauZtSYSibQVns9ikRL6z01kuwwRyQGphHgA2AI86Jy7ARhiEVMnzrntzrkW51xLPB5fYpmFpSISoF8jcRFJQSohfhI46Zzb6T1/hGSod5pZA4B37MpMiYUnFtZ0ioikZsEQd86dBk6Y2SavaRuwF3gcuM9ruw/4aUYqLECaExeRVAVSPO/TwHfNLAgcBj5K8hfAj8zsY8Bx4EOZKbHwVERKdLOPiKQkpRB3zr0CtMzx0ra0ViNAMsSHxyYZn5yipFj3Y4nI5SkhfKiiVHdtikhqFOI+pOVoRSRVCnEf0vopIpIqhbgPxSLJjyoGRnTDj4jMTyHuQ9PTKbrhR0QWohD3Ia0pLiKpUoj7UExz4iKSIoW4D4UCRQSLi7QIlogsSCHuQ2ZGLBLQdIqILEgh7lOxcIk+2BSRBSnEfSoaKaFflxiKyAIU4j4VC2tNcRFZmELcp7SSoYikQiHuU9qiTURSoRD3qekPNp1z2S5FRHxMIe5TsUiAsckpRiemsl2KiPiYQtyntH6KiKRCIe5TMa2fIiIpSCnEzeyomb1qZq+YWavX9nkza/faXjGzOzNbamGJhZPL0Z7Vh5siMo9UN0oGeKdzrvuitq84576UzoIkSSNxEUmFplN8amY5Ws2Ji8g8Ug1xBzxlZi+Z2f2z2j9lZrvN7JtmVpWB+gqWPtgUkVSkGuJvc85tAd4D/KmZ3QI8CFwBbAY6gC/P9UYzu9/MWs2sNZFIpKHkwhD15sS1foqIzCelEHfOnfKOXcBjwI3OuU7n3KRzbgr4BnDjZd673TnX4pxricfj6ao774VLigkFijQSF5F5LRjiZlZmZtHpx8DtwB4za5h12t3AnsyUWLhiWj9FRBaQytUp9cBjZjZ9/veccz83s38ws80k58uPAp/MVJGFKrmSoaZTROTyFgxx59xh4Po52u/NSEUyQyNxEVmILjH0sYqIdvcRkfkpxH0sFi7RjvciMi+FuI8lN0vWnLiIXJ5C3Me0priILEQh7mOxSAkTU45z45PZLkVEfEoh7mPnb73XlIqIzE0h7mOxyPSt9/pwU0TmphD3MS2CJSILUYj72PRytH3DCnERmZtC3Meqy4IA9A6PZbkSEfErhbiPzYT4kEJcROamEPex0mByOVqFuIhcjkLcx8yMmrKgQlxELksh7nNVCnERmYdC3Oeqy4L0KMRF5DIU4j6XnE4ZzXYZIuJTCnGfqyoLcmZI14mLyNwU4j5XUxZkcHSC0QktgiUil0opxM3sqJm9amavmFmr11ZtZk+bWZt3rMpsqYWpuiwE6FpxEZnbYkbi73TObXbOtXjPHwB2OOc2Aju855JmuuFHROaznOmU9wMPe48fBj6w7GrkEgpxEZlPqiHugKfM7CUzu99rq3fOdQB4x7pMFFjoasqTId4zqBAXkUsFUjzvbc65U2ZWBzxtZvtT/QFe6N8PsHbt2iWUWNgaKsIAnDp7LsuViIgfpTQSd86d8o5dwGPAjUCnmTUAeMeuy7x3u3OuxTnXEo/H01N1ASkNBqgqLaH9jEJcRC61YIibWZmZRacfA7cDe4DHgfu80+4DfpqpIgvd6soIp/oU4iJyqVSmU+qBx8xs+vzvOed+bmYvAj8ys48Bx4EPZa7Mwra6MsLxnuFslyEiPrRgiDvnDgPXz9HeA2zLRFFyocbKCM8d6sE5h/fLVEQE0B2bOWFNdSmDoxNaCEtELqEQzwFX1ZcD8HrnQJYrERG/UYjngE31UQBeP60QF5ELKcRzQDwaorK0hL0d/dkuRUR8RiGeA8yMmzZU85uDyQ83RUSmKcRzxK2b6mjvO8crJ/qyXYqI+IhCPEf8m+tXUxEp4T89spvnD/cwNaURuYikvnaKZFl5KMCD/24Lf/aDXdyz/XmaqiJsXVdFU1WE2vIQ8WiI2vLQzONYOKBrykUKgK3kHGtLS4trbW1dsZ+Xj4bHJnjytdP8bHcH+zoGON0/wuQco/JgcRG15cFLwr22PEhtNERNWYh4NEhNWYiKSAlFRQp8Eb8ys5dm7eVwAY3Ec0xpMMDdNzRx9w1NAExNOc4Mj9E9OEb34CiJgdHkcebxGKfOjrC7/Sw9g6PMNQsTKDKqy4LUloeoKQ8S947J517we78IqsuCBAOahRPxC4V4jisqMmq8sN1EdN5zJ73A7/ECP/k1Ro/3eLr9cGKI7sFRRiem5vw+FZGSmZCfDviastClbeVBykOa1hHJJIV4ASkuspkR9UKB75xjaGyS7oFReoZGSQyM0TM0Svf00fsFsP/0AD2DPZw9Nz7n9wkFii4I9ulpnelpnukpnng0pMAXWQKFuMzJzCgPBSgPBVhfW7bg+WMTU/QOnR/hT4/qe4bG6B5ITu90LDCtEwoUzRHwXujPmt+PR0OUBvVXVwQU4pImwUARqyrCrPJ2IprP9LTOzPz9Bcdk+4neYV4+dobe4THm+uy9LFhMbTREvPzSwJ8e2U+3hUuKM9BjEX9QiMuKmz2ts5CJyeQIv2uOoJ9+figxyPNHeugbnntKJxoOJMPeC/3pgI9HQ9RFQ9THwtRFQ1SVBnWVjuQchbj4WqC4iLpYmLrYwiP8sYmpmXn7xOCId0wG/fRx3+l+ft02ysDIxCXvLyk24uUh6mJh6mMh6qLnj3WznivsxU8U4pI3goEiGioiNFREgIp5zx0ZnyQxMErXwAid/aN09o/QNZA8JgZGOdI9xM4jvXOO7gNFRl00RDwWpj4aoi4Won466GPhmcfVCntZAQpxKUjhkmLWVJeyprp03vMuDvuu/hE6B0bp6k+2HesZ5oWjlw/7eDTEqoowqysjNFZGWO09Xl0ZoakqQkWkRFfkyLKkHOJmVgy0Au3OufeZ2eeBTwAJ75S/cs79U/pLFMmepYR91wUj+1FO959j76l+nt7bydhF196XBotnQr2xMszqigiNVZGZ0K+PhXVzlcxrMSPxzwD7gNistq84576U3pJEck8qYe+co2dojFN952g/c472vnOc6hvhVN85Tp09x95TZ+kevHALPjOoi4bOj969Y/IrTGOlRvOFLqUQN7Mm4L3A3wJ/kdGKRPKU2fmrct7UVDnnOSPjk8lQ98I9GfTTIT//aL5x9oi+8vxoflVFmJJijebzVaoj8a8CfwmX3Ob3KTP7Q5LTLP/BOXcmjbWJFJxwSTHN8XKa4+Vzvj41dX40fz7kzwf+nvazl2yobQb10TCrvXBvqiplXc30VxkNsbA+gM1hC65iaGbvA+50zv2Jmd0K/EdvTrwe6AYc8DdAg3Puj+d4//3A/QBr167deuzYsfT2QEQuMNdovv2C0D/H+OT5/++DgSLWVEVYX1PG2prSC46NlRHNyfvAfKsYphLi/wO4F5gAwiTnxB91zn1k1jnrgSecc9fN9720FK1I9k1OOU71neN47zBHe4Y43pM8HusZ5ljPMOfGJ2fOLTJorIqwrrrsgtH7htoy1teUKeBXyLJC/KJvdCvnR+INzrkOr/2zwE3OuXvme79CXMTfnHMkBkdnAv3YTLgPcax3+IJLKYuLjPU1pVxVH2VjXTlXescNtWVa6iDNMrWe+BfNbDPJ6ZSjwCeX8b1ExAfMLHmHajTMm9dXX/L62eFxjvUOcSgxyMGuQdo6BzlweoAnXzs9s6hZkcG6mjI21pWzsb6cjXVRrqwr58q6coV7BmhnHxFZtpHxSY50D9HWNcjBzgHaugZp6xrkaPcQE166m8Ha6tLkqL0uOhPyV9aVa1XKBWhnHxHJqHBJMVc3xLi6IXZB+9jEFEd7hmjrHKSta8AL+UF+9Xrigg9Xm6oivGFVjOsaY1y3uoI3NlVQFw3p+vcUKMRFJGOCgSKuqo9yVX0UaJhpH5+c4ljPMAe7BmjrHOT1rkH2dfSzY3/nzNLDteWhmVC/rrGCLWsrU1oIrdAoxEVkxZUUF83Mk98x65q2odEJ9nX0s6f9LHtOJY/PtnXPbAbeWBlh89pKblhTyQ1rq7h2dazg59kV4iLiG2WhAC3rq2mZ9aHqyPgkr53q55UTfbx8/AyvHO/jZ7s7AAgWF3HN6hg3rE2G+o3rq1PamCSf6INNEck5nf0j7Drex64TZ9h1vI/dJ/sYGU8uR7CuppSbNlTzluYabmquobEykuVqly9t14kvl0JcRDJhfHKK/R0D7DzSw/OHe3nxaO/M5t1rq0t5x8Za3rExzluvrCEWLslytYunEBeRgjI15dh/OhnqvznYw3OHuhkam6S4yNi8ppJbNsa5dVOcNzZW5MS6MQpxESloYxNT7Dp+hmfbunm2LcHu9rM4B6tiYW67pp53X7uKm5qrfbvao0JcRGSW3qExfrm/i6f2nuZXrycYGZ8iFg6w7ep6br+mnt/ZFPfVDUgKcRGRyzg3NsmzbQmefK2THfs76RseJxQo4h0ba7lrcyO3X1Of9csYdcemiMhlRILF3H7tKm6/dhUTk1O8cLSXp17r5MnXTvMv+7qIhgLc+cYG7t7SyI3rq303h66RuIjIHKamHM8f7uHRXe3886sdDI1N0lgZ4fe2NHL3DY2X3bgjEzSdIiKyDMNjEzy9t5OfvNzOv7YlmHJwc3MN9968jtuuqc/4B6IKcRGRNOnsH+GRl07yvZ3Hae87R100xD1vXsOHb1pLQ0VmbixSiIuIpNnklOOZA1185/ljPPN6giIztr2hjj9++wZu2lCd1hUY9cGmiEiaFRcZ266uZ9vV9ZzoHea7O4/zo9YTPLW3k+ubKvjELc3cce0qApmeatFIXEQkPUbGJ/nJyyd56NkjHOkeYk11hE+8o5nfb1mzrMsUNZ0iIrKCJqccT+/tZPuvD/Hy8T7i0RBf+4PNvPXK2iV9P02niIisoOIi447rVvHua+vZeaSXB585xIZ4WUZ+VsohbmbFQCvQ7u12Xw38EFhPcqPk33fOnclEkSIiucjMeEtzDW9prsnYz1jMjPtngH2znj8A7HDObQR2eM9FRGQFpRTiZtYEvBd4aFbz+4GHvccPAx9Ia2UiIrKgVEfiXwX+Epia1VbvnOsA8I51c73RzO43s1Yza00kEsupVURELrJgiJvZ+4Au59xLS/kBzrntzrkW51xLPB5fyrcQEZHLSOWDzbcBd5nZnUAYiJnZd4BOM2twznWYWQPQlclCRUTkUguOxJ1zn3PONTnn1gP3AL9wzn0EeBy4zzvtPuCnGatSRETmtJz7Qb8A3GZmbcBt3nMREVlBi7rZxzn3DPCM97gH2Jb+kkREJFUretu9mSWAY0t8ey3QncZycoH6XBjU58KwnD6vc87NeWXIiob4cphZ6+XWDshX6nNhUJ8LQ6b6nNk1EkVEJKMU4iIiOSyXQnx7tgvIAvW5MKjPhSEjfc6ZOXEREblULo3ERUTkIgpxEZEclhMhbmZ3mNkBMztoZjm7brmZrTGzX5rZPjN7zcw+47VXm9nTZtbmHatmvedzXr8PmNm7Z7VvNbNXvde+buncWjsDzKzYzHaZ2RPe87zus5lVmtkjZrbf++99cwH0+bPe3+s9ZvZ9MwvnW5/N7Jtm1mVme2a1pa2PZhYysx967TvNbP2CRTnnfP0FFAOHgGYgCPwWuCbbdS2xLw3AFu9xFHgduAb4IvCA1/4A8D+9x9d4/Q0BG7w/h2LvtReAmwED/hl4T7b7t0Df/wL4HvCE9zyv+0xyjf2Pe4+DQGU+9xloBI4AEe/5j4A/yrc+A7cAW4A9s9rS1kfgT4D/6z2+B/jhgjVl+w8lhT+0m4EnZz3/HPC5bNeVpr79lOS6MweABq+tATgwV1+BJ70/jwZg/6z2DwN/l+3+zNPPJpK7P72L8yGet30GYl6g2UXt+dznRuAEUE1yOY8ngNvzsc8kt6ScHeJp6+P0Od7jAMk7PG2+enJhOmX6L8e0k15bTvP+mXQDsJPLb7Bxub43eo8vbverr5L6piL50OdmIAF8y5tCesjMysjjPjvn2oEvAceBDuCsc+4p8rjPs6SzjzPvcc5NAGeBeTfozIUQn2s+LKevizSzcuAnwJ875/rnO3WONjdPu+8sYVORnO8zyRHUFuBB59wNwBDz70Gb83325oHfT3LaYDVQZmYfme8tc7TlVJ9TsJQ+Lrr/uRDiJ4E1s543AaeyVMuymVkJyQD/rnPuUa+509tYg4s22Lhc3096jy9u96PpTUWOAj8A3mWzNhWBvOzzSeCkc26n9/wRkqGez33+XeCIcy7hnBsHHgXeSn73eVo6+zjzHjMLABVA73w/PBdC/EVgo5ltMLMgycn+x7Nc05J4n0D/PbDPOfe/Zr10uQ02Hgfu8T6x3gBsBF7w/sk2YGZv8b7nH+LTTTnc4jcVyYc+nwZOmNkmr2kbsJc87jPJaZS3mFmpV+s2YB/53edp6ezj7O/1QZL/v8z/L5Fsf0iQ4gcJd5K8kuMQ8NfZrmcZ/Xg7yX8a7QZe8b7uJDnntQNo847Vs97z116/DzDrU3qgBdjjvfa/WeDDDz98Abdy/oPNvO4zsBlo9f5b/yNQVQB9/u/Afq/efyB5VUZe9Rn4Psk5/3GSo+aPpbOPJLfA/DFwkOQVLM0L1aTb7kVEclguTKeIiMhlKMRFRHKYQlxEJIcpxEVEcphCXEQkhynERURymEJcRCSH/X+gB41Bl5wUXgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plotting loss\n",
    "import matplotlib.pyplot as plt\n",
    "plt.plot(loss_array)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}