Linear Regression with PyTroch.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "c89e8d4e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import numpy as np\n",
    "from sklearn.datasets import make_regression\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import warnings as wr\n",
    "wr.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "04ffbe96",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating a dataset\n",
    "X_numpy, y_numpy = make_regression(n_samples=100, n_features=1, noise=20, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "5db98df2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Training set as Tensor\n",
    "X = torch.from_numpy(X_numpy.astype(np.float32))\n",
    "# Target set as Tensor\n",
    "y = torch.from_numpy(y_numpy.astype(np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "945452bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshaping the Tensor y\n",
    "y = y.view(y.shape[0], 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "76a1d92c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.float32, torch.float32)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.dtype, X.dtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6f8eb79d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sample size and the nnumber of features\n",
    "n_samples, n_features = X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "27b9611d",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_size = n_features\n",
    "output_size = 1 # Target varible/feature size\n",
    "\n",
    "# creates an instance of the nn.Linea which is nothing but Linear Regression\n",
    "model = nn.Linear(input_size, output_size) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "dc5103a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setting learning rate (alpha) \n",
    "learning_rate = 0.01\n",
    "\n",
    "# Loss function as MSE for regression task\n",
    "criterion = nn.MSELoss()\n",
    "\n",
    "# creates an instance of the stochastic gradient descent (SGD) optimizer\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5a3e5b82",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 10, loss = 4360.5356\n",
      "epoch: 20, loss = 3255.1062\n",
      "epoch: 30, loss = 2454.8708\n",
      "epoch: 40, loss = 1874.9723\n",
      "epoch: 50, loss = 1454.3417\n",
      "epoch: 60, loss = 1148.9679\n",
      "epoch: 70, loss = 927.0892\n",
      "epoch: 80, loss = 765.7556\n",
      "epoch: 90, loss = 648.3654\n",
      "epoch: 100, loss = 562.8956\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_iters = 100\n",
    "for epoch in range(n_iters):\n",
    "    \n",
    "    # Forward pass \n",
    "    y_pred = model(X)\n",
    "    \n",
    "    # Loss calculation\n",
    "    loss = criterion(y_pred, y)\n",
    "    \n",
    "    # Backward pass and updating the weights\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "\n",
    "    # Making the optimizer zero gradient before new step\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "    if (epoch+1) % 10 == 0:\n",
    "        print(f'epoch: {epoch+1}, loss = {loss.item():.4f}')\n",
    "\n",
    "# Plot\n",
    "predicted = model(X).detach().numpy()\n",
    "\n",
    "plt.plot(X_numpy, y_numpy, 'g*')\n",
    "plt.plot(X_numpy, predicted, 'r')\n",
    "plt.title('Best Fit Line of Our Model')\n",
    "plt.xlabel('X_numpy')\n",
    "plt.ylabel('y_numpy')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1e010002",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([70.8819], grad_fn=<AddBackward0>)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test = torch.tensor([1], dtype=torch.float32)  # Convert to Float data type\n",
    "yy = model(X_test)\n",
    "yy"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c78e1269",
   "metadata": {},
   "source": [
    "## Try with another dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ecb1b96a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>area</th>\n",
       "      <th>rent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>31500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2100</td>\n",
       "      <td>35000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2500</td>\n",
       "      <td>41050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2250</td>\n",
       "      <td>36100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3000</td>\n",
       "      <td>52100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   area   rent\n",
       "0  2000  31500\n",
       "1  2100  35000\n",
       "2  2500  41050\n",
       "3  2250  36100\n",
       "4  3000  52100"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "#Importing the Dataset\n",
    "df = pd.read_csv(\"data/Rent.csv\", sep=';')\n",
    "df.head() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "ec33e618",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60, 2)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "1cb902ac",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "X = df.drop('rent', axis =1)\n",
    "y = df[['rent']]\n",
    "\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X = scaler.fit_transform(X)\n",
    "y = scaler.fit_transform(y)\n",
    "\n",
    "# Convert DataFrame columns to NumPy arrays\n",
    "X = torch.from_numpy(X.astype(np.float32))\n",
    "y = torch.from_numpy(y.astype(np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "f4d28a1e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Reshaping the Tensor y\n",
    "y = y.view(y.shape[0], 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "928f8b94",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sample size and the nnumber of features\n",
    "n_samples, n_features = X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "0b3c6779",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_size = n_features # As we have only one feature varible\n",
    "output_size = 1 \n",
    "\n",
    "# Creates an instance of the nn.Linea which is nothing but Linear Regression\n",
    "model2 = nn.Linear(input_size, output_size) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "ad320286",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setting learning rate (alpha) \n",
    "learning_rate = 0.01\n",
    "\n",
    "# Loss function as MSE for regression task\n",
    "criterion = nn.MSELoss()\n",
    "\n",
    "# creates an instance of the stochastic gradient descent (SGD) optimizer\n",
    "optimizer = torch.optim.SGD(model2.parameters(), lr = learning_rate)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "089f36ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch: 10, loss = 0.0816\n",
      "epoch: 20, loss = 0.0710\n",
      "epoch: 30, loss = 0.0640\n",
      "epoch: 40, loss = 0.0592\n",
      "epoch: 50, loss = 0.0561\n",
      "epoch: 60, loss = 0.0540\n",
      "epoch: 70, loss = 0.0526\n",
      "epoch: 80, loss = 0.0516\n",
      "epoch: 90, loss = 0.0510\n",
      "epoch: 100, loss = 0.0506\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABESklEQVR4nO3de1xUdf7H8fcZkAFRUH4CaqIgXTXNwm5qqa1lrbVJ98JWTM11rc3aanV3C/1Z629X282stFoTFS3Xymrt4nbB3LRSJK0sTTdRExN0ExQVhDm/P5CJcQ7IZe68no8Hj+R7zpz5DKPNm8/5fs8xTNM0BQAAEORs/i4AAADAEwg1AAAgJBBqAABASCDUAACAkECoAQAAIYFQAwAAQgKhBgAAhARCDQAACAmEGgAAEBIINUALYxiGpkyZ4tXnyM7OlmEYKigo8OrzNNeiRYt09tlnq1WrVmrXrp2/ywl4U6ZMkWEYTXpsZmamkpOTPVsQcBJCDdAANR/Stb8SEhI0ePBgvfPOO1573iNHjmjKlClatWpVg/ZftWqVW501X7fddpvlY9auXaspU6bo4MGDDXqOzMxMtWnTpoGvIHBt2bJFmZmZSk1N1QsvvKDnn3/+lI9Zs2aN0tPTlZiYKLvdruTkZI0bN067du3yQcU/SU5OlmEYGjJkiOX2F154wfm+5+Xl+bQ2wJ/C/V0AEEz+93//VykpKTJNU/v27VN2drZ+/vOf65///KeuvfZajz/fkSNHNHXqVEnSoEGDGvy43/zmN7rwwgtdxmp+Sz569KjCw3/6p7927VpNnTpVmZmZHutW3Hnnnbrttttkt9s9cjxvWLVqlRwOh2bNmqXTTz/9lPvPnj1b9913n7p37657771XnTp10jfffKO///3vWrp0qd5++23169fPB5VXi4yMVG5urn744Qd17NjRZdvixYsVGRmpY8eO+aweIBAQaoBGuOaaa9S3b1/n96NHj1ZiYqJeeuklr4Saprrssst00003WW6LjIz0+vOHhYUpLCzM68/THEVFRZLUoCC3Zs0aTZw4UQMGDNC7776r1q1bO7eNHz9e/fv310033aTNmzerffv2za6tsrJSDodDERERde7Tv39/rV+/XkuXLtV9993nHP/+++/173//W+np6Xr11VebXQsQTDj9BDRDu3btFBUV5dL5kCSHw6Enn3xSPXv2VGRkpBITEzVu3Dj9+OOPLvvl5eVp6NCh6tChg6KiopSSkqK77rpLklRQUKD4+HhJ0tSpU52nE5o7H6b2MaZMmaKHHnpIkpSSkuJ8jubOhbGaU5OcnKxrr71WH3/8sS666CJFRkaqe/fuWrhwodvjDx48qIkTJyopKUl2u12nn366/vznP8vhcDTo+Z999ln17NlTdrtdnTt31oQJE1xOryUnJysrK0uSFB8ff8qf67Rp02QYhhYsWOASaCQpNTVVf/nLX7R3714999xzzvFBgwZZdtdOnltSUFAgwzA0c+ZMPfnkk0pNTZXdbtfXX39d72uMjIzUDTfcoCVLlriMv/TSS2rfvr2GDh1q+bgPP/xQl112maKjo9WuXTtdf/31+uabb9z2+/jjj3XhhRcqMjJSqampLq/tZDk5OUpLS1NUVJTi4uJ02223affu3fXWD3gDnRqgEUpKSrR//36ZpqmioiLNnj1bhw8f1ogRI1z2GzdunLKzszVq1Cj95je/0Y4dO/T000/r888/15o1a9SqVSsVFRXpqquuUnx8vCZNmqR27dqpoKBAr732mqTqD9s5c+Zo/PjxSk9P1w033CBJ6t279ynrPHTokPbv3+8yFhcXJ5vN9feYG264Qd9++61eeukl/e1vf1OHDh2cz+0N27dv10033aTRo0dr5MiRevHFF5WZmam0tDT17NlTUvUpt4EDB2rPnj0aN26cunbtqrVr12ry5Mnau3evnnzyyXqfY8qUKZo6daqGDBmi8ePHa+vWrZozZ47Wr1/v/Nk/+eSTWrhwoZYvX645c+aoTZs2df5cjxw5og8++ECXXXaZUlJSLPe59dZbdffdd2vFihWaNGlSk3428+fP17Fjx3T33XfLbrcrLi7ulI+54447dNVVV+k///mPUlNTJUlLlizRTTfdpFatWrnt//777+uaa65R9+7dNWXKFB09elSzZ89W//79lZ+f7wxbX375pfPv5pQpU1RZWamsrCwlJia6HfPxxx/XI488oltuuUVjxoxRcXGxZs+ercsvv1yff/45E7DhWyaAU5o/f74pye3Lbreb2dnZLvv++9//NiWZixcvdhl/9913XcaXL19uSjLXr19f5/MWFxebksysrKwG1Zmbm2tZpyRzx44dpmmabsebMWOGy/ZTGTlypBkdHV3vPjU/r9rH7NatmynJXL16tXOsqKjItNvt5m9/+1vn2LRp08zo6Gjz22+/dTnmpEmTzLCwMHPXrl11Pm9RUZEZERFhXnXVVWZVVZVz/OmnnzYlmS+++KJzLCsry5RkFhcX1/taNm7caEoy77vvvnr36927txkXF+f8fuDAgebAgQPd9hs5cqTZrVs35/c7duwwJZkxMTFmUVFRvc9Ro1u3buawYcPMyspKs2PHjua0adNM0zTNr7/+2pRkfvTRR873oPbfrz59+pgJCQnmgQMHnGObNm0ybTab+ctf/tI5Nnz4cDMyMtLcuXOnc+zrr782w8LCzNofGwUFBWZYWJj5+OOPu9T35ZdfmuHh4S7jJ79uwBs4/QQ0wjPPPKP33ntP7733nnJycjR48GCNGTPG2V2RpGXLlik2NlZXXnml9u/f7/xKS0tTmzZtlJubK+mnuRwrVqzQ8ePHPVrno48+6qyz5uvkyaT+0KNHD1122WXO7+Pj43XWWWfpu+++c44tW7ZMl112mdq3b+/y8xsyZIiqqqq0evXqOo///vvvq6KiQhMnTnTpSo0dO1YxMTF66623Gl3zoUOHJElt27atd7+2bduqtLS00cevceONNza6QxYWFqZbbrlFL730kqTqCcJJSUkuP+Mae/fu1caNG5WZmenSBerdu7euvPJKvf3225KkqqoqrVy5UsOHD1fXrl2d+51zzjlup7Ree+01ORwO3XLLLS7vVceOHXXGGWc4/64DvsLpJ6ARLrroIpeJwrfffrvOP/983XPPPbr22msVERGhbdu2qaSkRAkJCZbHqJmgOnDgQN14442aOnWq/va3v2nQoEEaPny47rjjjmavGurVq1edy339qfaHZI327du7zDXatm2bvvjiizo/4Gt+flZ27twpSTrrrLNcxiMiItS9e3fn9saoCTM14aYuhw4dOmXwqU9dp7ZO5Y477tBTTz2lTZs2acmSJbrtttssryVT189Gqg4sK1euVFlZmQ4dOqSjR4/qjDPOcNvvrLPOcoYfqfq9Mk3Tcl9JlqfAAG8i1ADNYLPZNHjwYM2aNUvbtm1Tz5495XA4lJCQoMWLF1s+pubD2jAMvfLKK/r000/1z3/+UytXrtRdd92lJ554Qp9++mlIXAvmZHWtiDJN0/lnh8OhK6+8Ug8//LDlvmeeeaZXaqvL6aefrvDwcH3xxRd17lNeXq6tW7e6BF7DMFxeV42qqirLY0RFRTWpvosvvlipqamaOHGiduzYoTvuuKNJx2kKh8MhwzD0zjvvWL63ofh3GIGNUAM0U2VlpSTp8OHDkqpXw7z//vvq379/gz6oLrnkEl1yySV6/PHHtWTJEmVkZOjll1/WmDFjmnz11sbwxXM0Rmpqqg4fPtykTlO3bt0kSVu3blX37t2d4xUVFdqxY0eTjhkdHa3Bgwfrww8/1M6dO53PUds//vEPlZeXuyzrb9++vctptRpN6Radyu23367HHntM55xzjvr06WO5T+2fzcm2bNmiDh06KDo6WpGRkYqKitK2bdvc9jv5sampqTJNUykpKT4Pm4AV5tQAzXD8+HH961//UkREhM455xxJ0i233KKqqipNmzbNbf/Kykrn0uIff/zR7Tf5mg+k8vJySXIuH27o1X6bIjo62uvP0Ri33HKLPvnkE61cudJt28GDB50h0sqQIUMUERGhp556yuVnO2/ePJWUlGjYsGFNqumPf/yjTNNUZmamjh496rJtx44devjhh9WpUyeNGzfOOZ6amqotW7aouLjYObZp0yatWbOmSTXUZ8yYMcrKytITTzxR5z6dOnVSnz59tGDBApf3+quvvtK//vUv/fznP5dU3U0bOnSoXn/9dZcrJX/zzTdu78kNN9ygsLAwTZ061e3vsmmaOnDggAdeHdBwdGqARnjnnXe0ZcsWSdVzO5YsWaJt27Zp0qRJiomJkVQ9V2bcuHGaPn26Nm7cqKuuukqtWrXStm3btGzZMs2aNUs33XSTFixYoGeffVbp6elKTU3VoUOH9MILLygmJsb5ARMVFaUePXpo6dKlOvPMMxUXF6dzzz1X5557rsdeU1pamiTpD3/4g2677Ta1atVK1113nTPsWDl+/Lgee+wxt/G4uDj9+te/blY9Dz30kN58801de+21zuXeZWVl+vLLL/XKK6+ooKDAufT8ZPHx8Zo8ebKmTp2qq6++Wr/4xS+0detWPfvss7rwwgvdlt431OWXX66ZM2fqgQceUO/evZWZmalOnTppy5YteuGFF+RwOPT222+7XHjvrrvu0l//+lcNHTpUo0ePVlFRkebOnauePXs2a0KxlW7dujXo+kUzZszQNddco0svvVSjR492LumOjY11efzUqVP17rvv6rLLLtOvf/1rVVZWavbs2erZs6fLabjU1FQ99thjmjx5sgoKCjR8+HC1bdtWO3bs0PLly3X33XfrwQcf9OhrBerlt3VXQBCxWtIdGRlp9unTx5wzZ47pcDjcHvP888+baWlpZlRUlNm2bVuzV69e5sMPP2wWFhaapmma+fn55u2332527drVtNvtZkJCgnnttdeaeXl5LsdZu3atmZaWZkZERJxyeXfNku5ly5bVuY/VMaZNm2aedtppps1mO+Xy7pEjR9a5bDw1NdXl53Xyku5hw4a5Hc9q6fOhQ4fMyZMnm6effroZERFhdujQwezXr585c+ZMs6Kios7aajz99NPm2WefbbZq1cpMTEw0x48fb/74448u+zR0SXdtq1evNq+//nqzQ4cOZqtWrcyuXbuaY8eONQsKCiz3z8nJMbt3725GRESYffr0MVeuXFnnku4ZM2Y0uI66fpa1WS3pNk3TfP/9983+/fubUVFRZkxMjHndddeZX3/9tdvjP/roI+ffu+7du5tz5851/sxO9uqrr5oDBgwwo6OjzejoaPPss882J0yYYG7dutW5D0u64QuGaVrMZAMAAAgyzKkBAAAhgVADAABCAqEGAACEBEINAAAICYQaAAAQEgg1AAAgJLSoi+85HA4VFhaqbdu2AXdpeAAAYM00TR06dEidO3eWzVZ3P6ZFhZrCwkIlJSX5uwwAANAEu3fvVpcuXerc3qJCTdu2bSVV/1BqLmkPAAACW2lpqZKSkpyf43UJmlAzZ84czZkzRwUFBZKknj176tFHH9U111zT4GPUnHKKiYkh1AAAEGRONXUkaCYKd+nSRf/3f/+nDRs2KC8vT1dccYWuv/56bd682d+lAQCAABDU936Ki4vTjBkzNHr06AbtX1paqtjYWJWUlNCpAQAgSDT08ztoTj/VVlVVpWXLlqmsrEyXXnppnfuVl5ervLzc+X1paakvygMAAH4QNKefJOnLL79UmzZtZLfb9atf/UrLly9Xjx496tx/+vTpio2NdX6x8gkAgNAVVKefKioqtGvXLpWUlOiVV17R3//+d3300Ud1BhurTk1SUhKnnwAACCINPf0UVKHmZEOGDFFqaqqee+65Bu3PnBoAAIJPQz+/g+r008kcDodLJwYAALRcQTNRePLkybrmmmvUtWtXHTp0SEuWLNGqVau0cuVKf5cGAAACQNCEmqKiIv3yl7/U3r17FRsbq969e2vlypW68sor/V0aAAAIAEETaubNm+fvEgAAQAAL6jk1AAAgMOQV5umKBVcorzDPbzUQagAAQLMt3LRQuQW5WrRpkd9qCJrTTwAAILDsPLhT+4/sl2EYWrp5qSTp5c0va2SfkTJNUx1ad1C3dt18Vg+hBgAANEnyrGS3saKyIqU9n+b83szy3eXwOP0EAACaJCc9R+E26/5IuC1cOek5Pq2HUAMAAJpkQNcByr4+23Jb9vXZGtB1gE/r4fQTAABoEqvTTzVGLB8hidNPAAAgCOSk5yjMCLPcFmaE+fz0E50aAADQJBm9M3RO/DkuE4NrrBu7Thd0usCn9dCpAQAAzWY7ESlsfowWhBoAANBkCdEJ6timo9I6p2nusLlK65ymjm06KiE6wee1GKZp+m4Gj5+VlpYqNjZWJSUliomJ8Xc5AACEhPLKckWERcgwDJmmqYqqCtnD7R47fkM/v5lTAwAAmqV2gDEMw6OBpjE4/QQAAEICoQYAAIQEQg0AAAgJhBoAABASCDUAACAkEGoAAEBIINQAAICQQKgBAAAhgVADAABCAqEGAACEBEINAAAICYQaAAAQEgg1AAAgJBBqAABASCDUAACAkECoAQAAIYFQAwAAQgKhBgAAhARCDQAACAmEGgAAEBIINQAAICQQagAAQEgg1AAAgJBAqAEAACGBUAMAAEICoQYAAIQEQg0AAAgJhBoAABASCDUAACAkEGoAAEBIINQAAICQQKgBAAAhgVADAABCAqEGAACEBEINAAAICYQaAAAQEgg1AAAgJBBqAABA8z32mDRhgmSafish3G/PDAAAgt9330mpqT9936+flJHhl1IINQAAoGkMw33sppt8X8cJnH4CAACN88EH1oHGNCW73ff1nECnBgAANIxpSjaLfsiaNdWnnfyMTg0AADi1p55yDzSxsdVBJwACjUSnBgAA1Ke8XIqMdB/ftUtKSvJ9PfWgUwMAAKxlZLgHmvT06u5MgAUaiU4NAAA42d69UufO7uNHjkhRUb6vp4GCplMzffp0XXjhhWrbtq0SEhI0fPhwbd261d9lAQAQWrp0cQ80f/5zdXcmgAONFESh5qOPPtKECRP06aef6r333tPx48d11VVXqayszN+lAQAQ/Navr16mvWeP67jDIT38sH9qaiTDNP14PeNmKC4uVkJCgj766CNdfvnlDXpMaWmpYmNjVVJSopiYGC9XCABAkLC65szbb0vXXOP7Wiw09PM7aDo1JyspKZEkxcXF+bkSAACC1OLFdV9EL0ACTWME5URhh8OhiRMnqn///jr33HPr3K+8vFzl5eXO70tLS31RHgAAga2qSgq3iABbtkhnneX7ejwkKDs1EyZM0FdffaWXX3653v2mT5+u2NhY51dSAC4/AwDApx580D3QXHxxdXcmiAONFIRzau655x698cYbWr16tVJSUurd16pTk5SUxJwaAEDLc/Cg1L69+/iPP0rt2vm6mkYJuTk1pmnqnnvu0fLly/Xhhx+eMtBIkt1uV0xMjMsXAAAtziWXuAeaBx+s7s4EeKBpjKCZUzNhwgQtWbJEb7zxhtq2basffvhBkhQbG6uoAF83DwCAX2zZIp1zjvt4ZaUUFub7erwsaE4/GVazsyXNnz9fmZmZDToGS7oBAC2G1edmTk71rQ+CTEM/v4OmUxMk2QsAAP96+21p2DD38RbwORo0oQYAANTDNCWbxVTZ9eulvn19X48fBM1EYQAAUIf/+z/3QNOlS3XQaSGBRqJTAwBA8Dp6VGrd2n18716pY0ff1+NndGoAAAhG6enugWbEiOruTAsMNBKdGgAAgsvu3VLXru7jx45Jdrvv6wkgdGoAAAgWMTHugeapp6q7My080Eh0agAACHxr10r9+7uPOxzW16NpoejUAAAQyAzDPdB88EF1d4ZA44JQAwBAIJo3zzq0mKZ0xRW+rycIcPoJAIBAcvy4FBHhPv6f/0jdu/u+niBCpwYAgEAxfrx7oPnZz6q7MwSaU6JTAwCAv+3fL8XHu4+Xlkpt2/q+niBFpwYAAH/q2dM90DzySHV3hkDTKHRqAADwhy+/lHr3dh+vqrK+MSVOiZ8aAAC+ZhjugeaVV+q+0zYahE4NAAC+8tpr0o03uo+bpu9rCUGEGgAAvM3hkMLC3Mc3bbI+BYUmoccFAIA3TZniHmjOPru6O0Og8Sg6NQAAeMPhw9arl4qLpQ4dfF9PC0CnBgAATxs61D3QjBtX3Z0h0HgNnRoAADxlxw7rK/9WVEitWvm+nhaGTg0AAJ5gs7kHmhdeqO7OEGh8gk4NAADNkZtrfddslmn7HKEGAICmqOtCeR9/LPXv7/t6wOknAAAa7emn3QNNdHR10CHQ+A2dGgAAGqq8XIqMdB/fuVPq2tX39cAFnRoAABpi5Ej3QPOLX1R3Zwg0AYFODQAA9dm3T+rY0X28rExq3dr39aBOdGoAAKhLt27ugeZPf6ruzhBoAg6dGgAATpafL6WluY87HJJh+L4eNAidGgAAajMM90CzYkV1d4ZAE9Do1AAAINUdWLiIXtCgUwMAaNmOHbMMNEPua6/8wg3aULhBOw/u9ENhaCw6NQCAlquO7owxRTJ0UGnP/3QaysyiYxPo6NQAAFqedessA81lo6oDjSSZqg4x4bZw5aTn+LA4NBWdGgBAy1JHdya/cIM+ft59xdNnYz7TBZ0u8HZV8AA6NQCAluGhh6wDzeHDLpOBDRku/0XwoFMDAAh9DVjZlBCdoI5tOsphOlRUVqT46HjZDJsSohN8VCSai04NACB0GYZ1oDFNl0Cz8+BO7Tu8T8tvXS7zxLhpmlp+63LtO7yP1U9Bgk4NACD0mKZks/i9vV8/ac0at+HkWcnOP9ecdtp/ZL8unXfpT4dk9VPAo1MDAAhqeYV5umLBFcorzKseMAzrQGOaloFGknLScxRuq/49v2bVE6ufgg+hBgAQ1GaunancglzNf2Oq9ammRYtOeVXgjN4Z+mzMZ5bbPhvzmTJ6Z3iiVHgZp58AAEFn58Gd2n9kvwzD0Jtb35Q5RZJWuO/YhFsc2GSTQw7nfxE8CDUAgKBTMwfmkVXSkVXu2097QNrzROMCTc3qp6SYJI0+f7TmfT5Pu0t3s/opiBBqAABBqbo7486oY/xUusR0UcF9Bfpi3xf63fu/0zM/f0a9E3vLHm5vaonwMUINACC4GIasejBGliRDCjPCtGD4giYd2h5u1x8+/INyC3L1hw//oH/d+a9mlQrfItQAAIJHPTegrLEgfYEyejVuYm/tOTof7vhQkpS7I1f5e/NlmqY6tO6gbu26NbVq+AihBgAQ+BoQZmrERcY1+vC1r1NTo9Ks5C7dQYYl3QCAU3K7Foyv/PijdaC5/XZ9X7JbidGJ6tupr+YOm6u+nfoqMTpRvRJ7+bZGBAw6NQCAU1q4aaFyC3K1aNMi9e3c1zdPeor7NXWRtHPiTkWERcgwDN2ddrcqqiqaNLF3fN/xmpM3p97tCHx0agAAlnYe3KkNhRuUvzdfSzcvlSS9vPll5e/N14bCDd67H9KyZdaBZvVqt+vO2MPtMk7saxhGk1cq/a7/7+oMLuP7jtfv+v+uSceFbxmm2YQrEwWp0tJSxcbGqqSkRDExMf4uBwACmjH1p2BhyJAp0/nfGh6fZ9KAu2l7Q+3XWhfm1PhPQz+/6dQAACz59H5IF11kHWiOHvV6oJGqX2uYEWa5LcwI495PQYJODQCgTvl7811WANXYcPcGXdDpAs88iZ+6MyfzyWtFk9CpAQB4jCHD5b+eOahhHWhM0+eBpjbbiY9GGx+RQYd3DABQp5r7IcVHx0uS4qPj1bFNx+bdD+n4cesw062bX8NMzWtN65ymucPmKq1zWvNfK3yK008AAEs1V9k97jiuX7z0CxUfKVZ863i9efubamVr1bSr7AbIqaa6lFeWO5eIm6bZ5CXi8KyGfn5znRoAgKXaV9mtOe20/8h+XTrvUud4g1cE5edLae7zVfTii9KoUc0p06NqB5jmLBGHfxBqAACWctJzlPlGpiodlZarn7Kvz27YgQK8O4PQwZwaAICljN4Z+mzMZ5bbPhvzmTJ6n+Kmkffeax1o9u4l0MArgirUrF69Wtddd506d+4swzD0+uuv+7skAGgRGr0iyDCkp592HzdNqWNHD1YG/CSoQk1ZWZnOO+88PfPMM/4uBQBahEavCArQZdpoGYJ29ZNhGFq+fLmGDx/e4Mew+gkAGq/BK4KYOwMv4eJ7ksrLy1VaWuryBQD4SV5hnq5YcIXyCvPq3OeUN42kO4MAEdKhZvr06YqNjXV+JSUl+bskAAgoCzctVG5BrhZtWtT4BxcWWoeZ++8nzMAvQvr0U3l5ucrLy53fl5aWKikpidNPAFq0movqGYahaxZfo6KyIiVEJ+idjHdkmmbDLqrHqSb4EBffk2S322W3c+EkAC1PXmGeHn7vYf3lyr+ob+e+LttqX1SvRlFZkcvNHOu8qN7zz0vjxrmPb9wonXdeMyoGmi+kTz8BQEtV32mlnPQchdusf6cNt4UrJz3H+qCGYR1oTJNAg4AQVKHm8OHD2rhxozZu3ChJ2rFjhzZu3Khdu3b5tzAACAA7D+7UhsINyt+br6Wbl0qSXt78svL35mtD4QbtPLhTkjSg64A6rwacfX22BnQd4DrYpYv16abjxzndhIASVHNqVq1apcGDB7uNjxw5UtnZ2ad8PEu6AYQyY2od81xqMbPMBu9XfVDmzsD/QnJJ96BBg2SapttXQwINAIS6hp5WyknPUZgRZrlfmBFWvR/LtBGEgirUAADqltwuWVmXZ1luy7o8S8ntkiVV39Np3dh1lvutH7FaGeeNcN/Qrx9hBgEvpFc/AUBLMmD+gDq3PbLqEWmV+6omm2xyyCGbbKqa4pCm9Hd/MGEGQYJODQCEiPF9xzd4e+17Or3R8f7qQHOy114j0CCo0KkBgBDxu/6/U+e2nfVI7iNu26YNnqY7e9/p/L5LTBcV3Fcge6tISevdD0aYQRAKqtVPzcXqJwChrFGrmm6+WXrlFfcdfvxRatfOs4UBzRSSq58AAHXLSc+RrY7/rdtk++mieoZhHWhMk0CDoEaoAYAQkdE7Q+vvtjiVJGn93eurVzWxTBshjFADACGopmNjk00ypQs6p1nvSJhBCGGiMACEkJpVTUkxSRp9/miNu/BX1jsSZhCCCDUAEEJqVjVFfLdTxllnue/w+OPS73/v+8IAHyDUAECIqV6mbYHuDEIcc2oAIFTcdJP1ROBt2wg0aBHo1ABAKOBu2gCdGgAIanXdTbuqikCDFodODQAEK7ozgAtCDQAEG8IMYInTTwAQLIqLrQNNmzYEGkB0agAgONCdAU6pSZ2ahQsXqry83G28oqJCCxcubHZRAIAT/vpX60CzaBGBBjiJYZqN/1cRFhamvXv3KiEhwWX8wIEDSkhIUFVVlccK9KSG3rocAAIC3RlAUsM/v5vUqTFNU4bFP7bvv/9esbGxTTkkAKBGXcu0S0sJNEA9GjWn5vzzz5dhGDIMQz/72c8UHv7Tw6uqqrRjxw5dffXVHi8SAFoMujNAkzUq1AwfPlyStHHjRg0dOlRt2rRxbouIiFBycrJuvPFGjxYIAC0CYQZotkaFmqysLElScnKybr31VkVG1nHTNABAwxw/LkVEWG8j0ACN0qQl3SNHjpRUvdqpqKhIDofDZXvXrl2bXxkAhDq6M4BHNWmi8LZt23TZZZcpKipK3bp1U0pKilJSUpScnKyUlBRP1wgAoeX9960Dzb33EmiAZmhSpyYzM1Ph4eFasWKFOnXqZLkSCgBgge4M4DVNCjUbN27Uhg0bdPbZZ3u6HgAITRddJK1f7z7+7bfSGWf4vh4gBDUp1PTo0UP79+/3dC0AEJrozgA+0aQ5NX/+85/18MMPa9WqVTpw4IBKS0tdvgAAqvsieqZJoAG8oEm3SbDZqrPQyXNpaq40zG0SALR4dGcAj2no53eTTj/l5uY2uTAACGmEGcBvmhRqBg4c6Ok6ACC47dghde/uPt6zp/TVV76vB2iBmjSnRpL+/e9/a8SIEerXr5/27NkjSVq0aJE+/vhjjxUHAEHBMKwDjWkSaAAfalKoefXVVzV06FBFRUUpPz9f5eXlkqSSkhL96U9/8miBABCw7rrL+nTTW29xugnwgyaFmscee0xz587VCy+8oFatWjnH+/fvr/z8fI8VBwAByzCk+fPdx01T+vnPfV8PgKaFmq1bt+ryyy93G4+NjdXBgwebWxMABK66lmkfO0Z3BvCzJoWajh07avv27W7jH3/8sbpbnVcGgFBQx8qm25bdKtntPi4GwMmatPpp7Nixuu+++/Tiiy/KMAwVFhbqk08+0W9/+1s9+uijnq4RAPyrjjDT+rEoHa08qqitbyp/b75M01SH1h3UrV03HxcIQGpiqJk0aZIcDod+9rOf6ciRI7r88stlt9v10EMPacyYMZ6uEQD8o7RUio213GRMkVR5VJJ0tPKo0p5Pc24zszgNBfhDk04/GYahP/zhD/rvf/+rr776Sp9++qmKi4sVGxurlJQUT9cIAL5nGJaBxphyItAACDiNCjXl5eWaPHmy+vbtq/79++vtt99Wjx49tHnzZp111lmaNWuW7r//fm/VCgDe99RT1qeb7rpLMk3lpOcozAizfGiYEaac9BwvFwigLo06/fToo4/queee05AhQ7R27VrdfPPNGjVqlD799FM98cQTuvnmmxUWZv2PHQACXgNucZDRO0MypBGvjXDbbUH6AmX0yvBWdQBOoVGdmmXLlmnhwoV65ZVX9K9//UtVVVWqrKzUpk2bdNtttxFoAASnupZpf/ed5TLtuMg4y8N8vvdzT1cGoBEaFWq+//57paVVT4Y799xzZbfbdf/997vdrRsAgkZ93Zk65gj2SuylxOhE9e3UV3OHzVVUeJQk6Z1t73irSgAN0KjTT1VVVYqIiPjpweHhatOmjceLAgCva8bdtLvEdNFLN76k70u/l81m07HKY5Kkb/Z/o8VfLpbpMJXSPkX9u/b3ZMUATqFRocY0TWVmZsp+4iJTx44d069+9StFR0e77Pfaa695rkIA8CTTlGx1NKkbcUXgKxZe4f5wmS5zbVjaDfhWo0LNyJEjXb4fMcJ9ohwABKxmdGdONr7veM3Jm1PvdgC+ZZhmy7lZSWlpqWJjY1VSUqKYmBh/lwPAV/79b8nifnXq0EEqLm7yYRd/udhyFVTODTmsggI8qKGf3026ojAABA0PdmcABLYmXVEYAAJejx7Wgebttz0WaM6MO1M2w6bWrVrrjnPvUOtWrWUzbDoz7kyPHB9A49CpARB6fNSdufC0C/Xjwz+qTUQb2Ww2ORwOHa44rJhITm8D/kCoARA66gozlZWSly4OWjvA2Gw2Ag3gR4QaAKGBuTNAi0eoARDcCDMATmCiMIDgtHcvgQaACzo1AIIPYQaABTo1AILHQw9ZB5qsLAINADo1AIIE3RkApxB0nZpnnnlGycnJioyM1MUXX6x169b5uyQA3mQY1oGmuJhAA8BFUIWapUuX6oEHHlBWVpby8/N13nnnaejQoSoqKvJ3aQC8ob7uTIcOvq0FQMALqlDz17/+VWPHjtWoUaPUo0cPzZ07V61bt9aLL77o79IAeFJd3RnTpDsDoE5BE2oqKiq0YcMGDRkyxDlms9k0ZMgQffLJJ5aPKS8vV2lpqcsXgAB2/DhzZwA0WdCEmv3796uqqkqJiYku44mJifrhhx8sHzN9+nTFxsY6v5KSknxRKoCmMAwpIsJ9nO4MgAYKmlDTFJMnT1ZJSYnza/fu3f4uCcDJXn/dujvTrx9hBkCjBM2S7g4dOigsLEz79u1zGd+3b586duxo+Ri73S673e6L8gA0BaeaAHhQ0HRqIiIilJaWpg8++MA55nA49MEHH+jSSy/1Y2UAGm3QIOtAs24dgQZAkwVNp0aSHnjgAY0cOVJ9+/bVRRddpCeffFJlZWUaNWqUv0sD0FB0ZwB4SVCFmltvvVXFxcV69NFH9cMPP6hPnz5699133SYPAwhAdYSZ+966V7N+/pSPiwEQigzTbDm/HpWWlio2NlYlJSWKiYnxdzlAy1FHoDGmSAnRCXon4x2ZpqkOrTuoW7tuvq0NQMBr6Od3UHVqAASZesKMIUOSqeKyYqU9n+bcZma1mN+zAHhY0EwUBhBEdu2yDDSHuiSo1f9W/y5lynT5b7gtXDnpOb6rEUDIoVMDwLPqmQjcVtJne/NdOjM1PhvzmS7odIF3awMQ0ujUANDCTQvV/s/ttXDTwqYf5I9/tA40r7/OyiYAPkGnBoCe+OQJHTx2UH/95K/65Xm/bPwBGrFMOyE6QR3bdFTJsRIdrTyqqPAoxUbGKiE6ofHPCwC1sPoJaKHW7FqjHT/ukGEzNOK1Ec7xnBtyZDpMpbRPUf+u/es/SF1h5uhRKTKyzuesclRp1JujZMqUIUPzfzFfYbawhj0ngBanoZ/fhBqghTKm1hFIajl5JVJeYZ4efu9h/eXKv6jvaRfW8aC6/5fSlOcEgIZ+fjOnBmihRvQa0ejtM9bM0IeZudaBpgF30x7fd3yztgNAfejUAC1UQ7smOw/u1P4j+xV29Jj6nD6gjh0b/r+RxV8udjndVSPnhhxl9Mpo8HEAtBxcfA+ARyTPSpY5xXqbcWK8xfxmBCCgcfoJaKFy0nNkq+N/ATbZqi+El5NjGWimDPwp0DTWmXFnymbYFG6r/p0q3BYum2HTmXFnNu2AAHACnRqghcronSEZsjwVtPCGhcrobT3npnaYCTPCtGD4ggY/586DO2UzbModmasbl96o/Uf3q529nV699VXZDJt2HtzJvZ8ANBmhBmjB4iLj3MbMKZKmuAealPukgvauYwvSFzRqHkzyrGTnn6vv/SQdOHpAA7MH/vT8rH4C0EScfgJasF6JvZQYnajIsOprytQ1d+adb992CzSSdSiqT056jvO0E/d+AuBpdGqAFqrmQnh//tmfNfKCTOudTqxq6lX6vRKjE5UUm6Qx54/R3z//u3aX7FavxF6Nes6M3hk6J/4c7v0EwCsINUALNWD+AMmUzKnW240pP61q6hLTRTsn7lREWIQMw9DdaXeroqpC9nB7k5/fJpsccjj/CwDNRagBWqhTLdM++UJ4tQOMYRhNDjQ1935KiknS6PNHa97n87S7dDf3fgLQbFx8D2hp9uyRunRxG84aJP3voOo/Txs8TXf2vtNrK5HKK8udXR/TNJvd9QEQ2rhNAhBg8grzdMWCK5RXmOe/IgzDMtAYU34KNJL0SO4jLiuVPM0ebpdx4maYzen6AEBthBrAR2asmaHcglzNXDPT908+b57lHbXPuddmeRG9MCOMlUgAgg5zagAvqrlvkmEY+ue3/5Qkvfntm8rfmy/TNNWhdQfvX2zOIsxIkkxTi/fmW65EWjd2HSuRAAQdQg3gRVancI5WHnUJEl672NzZZ0tbt7qPV1VJNtcmrSFDpkznfwEgGBFqgFBUT3emtpqVSCXHSnS08qgiwyMVGxnLSiQAQYk5NYAX5aTnKMwIs9zmlXkrhmEdaEzTLdCs2bVGq3as0vSfTdfRyqOSqrtI0382Xat2rNKaXWs8WxsAeBmhBvCijN4ZWpBufcPHBekLqm8q2QRuK6nKyqzDzB13uIWZGgPmD9Cdr9+pUW+Mchkf9cYo3fn6ndUX5/NEbQDgI4QawMvquj9SY++bVNvCTQuVW5CrRZsWVYeZNm3cdzJNafHiOo9x8sX1Gru9LjPXzqxe5bXWD6u8ALRozKkBvKzmppHNvW9S7ZVUSzcv1cAd0qwpT7nv+MEH0hVXnPJ4v+v/O3Vu21mP5D7itq11eGsNO2OYNhRuaNAKrdq1vbn1TUnSm1t9vMoLQIvHFYUBH/DEFXSNqT+dXqrrFgd1nWo61fFO5VQrtBpyLK+t8gIQ8riiMBBAPHEF3Zz0HL3wT8My0MT+MUyLNzVu0nF9k5hrhNvCuQgfgKBBqAGCRMZ5IzRmg3u3w5gi5f56XaMnHQ/oOkALhltPYq6RfX22BnQ99YRhn6/yAgALzKkB6pBXmKeH33tYf7nyL+rbua//CqnjmjPGFMkmmyRHkw7bkHs7jVg+QtKpTx1l9M6QDGnEayPcti1IX6CMXk1b5QUAjUGnBqiDywojf3A4LAPNV53DddHzF2rusLlK65ymjm06NulieZ4+/eSNVV4A0BhMFAZqqb2K55rF16iorEgJ0Ql6J+Md367iqeeKwJ6YdFwjv457P9XYcPeGBt8D6vvS79X3+b5KiknSmAvG6O/5f9fu0t3KuztPXWLc7wwOAA3V0M9vTj8BtVidkikqK/LNvZok6T//kU4/3X18/nwpM1OSXAJMUycdn+zkez415R5QXWK6aOfEnc7AdXfa3c0KXADQWIQaoJac9BxlvpGpSkel27ZwW7iyr8/23pM38H5NnlRz76fE6ETtOLhDlY5KhdvCldIuRfvK9jX6tJY3AhcANBShBqhlQNcByr4+2zlBtraGrgRqtKeeku67z338+++l007z/PPV0iWmiwruK1BEWIQqqirUytZKxx3Hnd8TSgAEE+bUALX4/CJyfujOAECw4eJ7QBPkpOco3GbdwPTohei6dLEONA4HgQYAmojTT0AtGb0zdE78OZYrgj4b81mDVwLVi+4MAHgFnRqgDrYT/zxsnvpnYhjWgcY0CTQA4AGEGuAkNSuC0jqnNfsCd5Kkgwetw8ykSYQZAPAgJgoDFjx2gTtONQFAszFRGGiGZt9V+623rAPNhg0EGgDwEiYKA55GdwYA/IJODeApw4dbB5rycgINAPgAnRrAE+jOAIDfEWqA5iDMAEDA4PQT0BTHj1sHmqFDCTQA4Cd0aoDGojsDAAGJTg3QUF98YR1oli8n0ABAAKBTAzQE3RkACHh0aoD6TJliHWj27yfQAECAoVMD1IXuDAAEFTo1wMkGDeJu2gAQhOjUADVMU7JZ5Py4OOnAAd/XAwBoFEINIHGqCQBCAKef0LL997/Wgeallwg0ABBk6NSg5aI7AwAhhU4NWp7337cONLt2EWgAIIjRqUHLQncGAEJW0HRqHn/8cfXr10+tW7dWu3bt/F0Ogs2ECdaBpqqKQAMAISJoQk1FRYVuvvlmjR8/3t+lINgYhvTss65jSUl1L+E+Ia8wT1csuEJ5hXleLhAA4AlBc/pp6tSpkqTs7Gz/FoLg0cxTTQs3LVRuQa4WbVqkvp37erAwAIA3BE2oaYry8nKVl5c7vy8tLfVjNfCZ8nIpMtJ9/E9/kiZPrvehOw/u1P4j+2UYhnK+yJEkLfpikUb2GSnTNNWhdQd1a9fNG1UDAJoppEPN9OnTnR0etBDN7M4kz0p2G/vx2I9Kez7tp0NlMQcHAAKRX+fUTJo0SYZh1Pu1ZcuWJh9/8uTJKikpcX7t3r3bg9UjoHz9tXWgyc9v1ETgJ4c+qTAjzHJbmBGmJ4c+2cQCAQDeZpim/5Z+FBcX68Ap7qnTvXt3RUREOL/Pzs7WxIkTdfDgwUY/X2lpqWJjY1VSUqKYmJhGPx4ByoPLtI2pdRyr9mHp1ACATzX089uvp5/i4+MVHx/vzxIQzGbNkiZOdB8/ckSKivJZGXmFeXr4vYf1lyv/woRiAPCjoFnSvWvXLm3cuFG7du1SVVWVNm7cqI0bN+rw4cP+Lg3+YBjWgcY0mxVoxvet/5IBVttrr5ICAPiPX08/NUZmZqYWLFjgNp6bm6tBgwY16BicfgoB554rbd7sPu7Bv8aPrX5Mj+Q+4jY+bfA0/fHyP0pyXSV1zeJrVFRWpIToBL2T8Q6rpADAw4Li9FNjZGdnc42alqyuC+VlZkrz53v0qWLs1v9gao9brZIqKitilRQA+FHQnH5CC2YY1oHGND0eaCTp0i6XypChqPAo3XHuHYoKj5IhQ5d2udS5T056jsJt1r8ThNvClZOe4/G6AAD1C5rTT57A6acg88MPUqdO7uMrVkjDhnn1qUuPlapNRBvZbDY5HA4drjismMif/s7sPLhTH+/6WCOWj3B7bE56jgZ0HcDpJwDwkIZ+fhNqEJgC/G7aLP0GAN9p6Oc3p58QWN54wzrQ7NsXMIFG4vQTAASioJkojBYgwLsztWX0ztA58ee4TAyu8dmYz3RBpwv8UBUAtGx0auB/d9xhHWgcjoAMNCeznfhnZOOfEwD4FZ0a+JdVmOnbV1q/3ve1NFJCdII6tumopJgkjT5/tOZ9Pk+7S3crITrB36UBQIvERGH4RxCdaqpPeWW5IsIiZBiGTNNURVWF7OF2f5cFACGFicIITGVl1oHm2WeDLtBIkj3cLuPE6zEMg0ADAH7E6Sf4Toh0ZwAAgYlODbxv/XrrQPPNNwQaAIDH0KmBd9GdAQD4CJ0aeMfUqdaBprycQAMA8Ao6NfA8qzBjGNXXnQEAwEvo1MBzxoyxDjSmSaABAHgdnRo0n8MhhYW5DWdeL8WO+41m+aEkAEDLQ6hB89jtUkWF23DijAQVlRUpYfPLGtlnpEzTVIfWHdStXTc/FAkAaAkINWia4mIpwf12AD1+LX2TIKmsSJJUVFbkctNHM4tJwgAA72BODRrPMCwDzeJNOdrW0Tonh9vClZOe4+3KAAAtGKEGDVdQYD0R+PBhyTQ1oOsAZV+fbfnQ7OuzNaDrAK+WBwBo2Tj9hIaJiJCOH3cdO/98KT/f+W3yrOQ6Hz5i+QhJnH4CAHgPnRrU76OPqrszJwca03QJNJKUk56jMMN9FZQkhRlhnH4CAHgVocYD8grzdMWCK5RXmOfvUjzLMKRBg1yGtrz6nK7IHmz5WjN6Z2jd2HWWh1o3dp0yemd4o0oAACQRajxi5tqZyi3I1cy1M/1dimfMmeM+d8Zul0xTUxwfNui12k781bLxVwwA4CPMqWminQd3av+R/TIMQ29ufVOS9ObWN5W/Nz94r8lSUVEdXk7y/aaPta9DpIy9+ad8rQnRCerYpqOSYpI0+vzRmvf5PO0u3a2EaPfVUgAAeJJhmi3n7oKlpaWKjY1VSUmJYmJimnUsY2odd5+uJagmxY4eLb34ouvYsGHSihWNfq3lleWKCIuQYRgyTVMVVRWyh7uHJQAAGqKhn990alq6oiIpMdF9/PBhKTq6SYesHWAMwyDQAAB8ggkPTRQSK326d3cPNNOmVa9sqhVoQuK1AgBCHp2aJsronSEZ0ojXRrhtW5C+QBm9Anilz+efSxdc4D7ucFheXC+oXysAoMWgU9MMcZFxjRoPCIbhHmjeeKO6O2N1teATgvK1AgBaFEJNM/RK7KXE6ET17dRXc4fNVd9OfZUYnaheib38XZq7f/zDOrSYpvSLX5zy4UH1WgEALRKrn5op4Ff6OBxSmMV8mM2bpR49GnWogH+tAICQ1NDPbzo1zWQPt8s40QEJuJU+v/+9e6A577zq7kwjA40U4K8VANDiMVE4FJWWSrGx7uMHDkhxzIEBAIQmOjWhZuBA90Bz773V3RkCDQAghNGpCRXbtklnnuk+fvy4FM7bDAAIfXRqQoFhuAea7Ozq7gyBBgDQQvCJF8zee0+66ir38ZazoA0AACdCTTAyTclm0WT79FPp4ot9Xw8AAAGA00/B5m9/cw808fHVQYdAAwBowejUBItjx6SoKPfx77+XTjvN9/UAABBg6NQEg9tucw80t9xS3Z0h0AAAIIlOTWArLLQOLUePSpGRvq8HAIAARqcmUHXs6B5oZsyo7s4QaAAAcEOnJtB89pl0ySXu4w6H9V22AQCAJDo1gcUw3APNu+9Wd2cINAAA1ItQEwgWLLAOLaYpDR3q+3oAAAhCnH7yp8pKqVUr9/Fvv5XOOMP39QAAEMTo1PjLxInugWbAgOruDIEGAIBGo1Pja//9r/Q//+M+XlIixcT4vh4AAEIEnRpfuuAC90Dzu99Vd2cINAAANAudGl/45hupRw/38cpKKSzM9/UAABCC6NR4m2G4B5qXX67uzhBoAADwGDo13rJihXTdde7jpun7WgAAaAEINZ5mmpLNogG2YUP1nBoAAOAVnH7ypD/9yT3QJCdXBx0CDQAAXkWnxhOOHJGio93H9+2TEhJ8Xw8AAC0QnRpPODnQjBpV3Z0h0AAA4DN0ajwhLa16zowklZdLERH+rQcAgBYoKDo1BQUFGj16tFJSUhQVFaXU1FRlZWWpoqLC36VVy8ur7syYJoEGAAA/CYpOzZYtW+RwOPTcc8/p9NNP11dffaWxY8eqrKxMM2fO9Hd5AAAgABimGZwXTpkxY4bmzJmj7777rsGPKS0tVWxsrEpKShTDbQkAAAgKDf38DopOjZWSkhLFxcXVu095ebnKy8ud35eWlnq7LAAA4CdBMafmZNu3b9fs2bM1bty4evebPn26YmNjnV9JSUk+qhAAAPiaX0PNpEmTZBhGvV9btmxxecyePXt09dVX6+abb9bYsWPrPf7kyZNVUlLi/Nq9e7c3Xw4AAPAjv86pKS4u1oEDB+rdp3v37oo4saKosLBQgwYN0iWXXKLs7GzZrG5HUA/m1AAAEHyCYk5NfHy84uPjG7Tvnj17NHjwYKWlpWn+/PmNDjQAACC0BcVE4T179mjQoEHq1q2bZs6cqeLiYue2jh07+rEyAAAQKIIi1Lz33nvavn27tm/fri5durhsC9IV6QAAwMOC4hxOZmamTNO0/AIAAJCCJNQAAACcCqEGAACEBEINAAAICYQaAAAQEoJi9ZOn1Ews5h5QAAAEj5rP7VMtEGpRoebQoUOSxD2gAAAIQocOHVJsbGyd2/16mwRfczgcKiwsVNu2bWUYhteep7S0VElJSdq9eze3YwhQvEeBj/co8PEeBbZQen9M09ShQ4fUuXPneu8o0KI6NTabze3ifd4UExMT9H+RQh3vUeDjPQp8vEeBLVTen/o6NDWYKAwAAEICoQYAAIQEQo0X2O12ZWVlyW63+7sU1IH3KPDxHgU+3qPA1hLfnxY1URgAAIQuOjUAACAkEGoAAEBIINQAAICQQKgBAAAhgVDjZQUFBRo9erRSUlIUFRWl1NRUZWVlqaKiwt+l4YTHH39c/fr1U+vWrdWuXTt/lwNJzzzzjJKTkxUZGamLL75Y69at83dJqGX16tW67rrr1LlzZxmGoddff93fJaGW6dOn68ILL1Tbtm2VkJCg4cOHa+vWrf4uyycINV62ZcsWORwOPffcc9q8ebP+9re/ae7cufr973/v79JwQkVFhW6++WaNHz/e36VA0tKlS/XAAw8oKytL+fn5Ou+88zR06FAVFRX5uzScUFZWpvPOO0/PPPOMv0uBhY8++kgTJkzQp59+qvfee0/Hjx/XVVddpbKyMn+X5nUs6faDGTNmaM6cOfruu+/8XQpqyc7O1sSJE3Xw4EF/l9KiXXzxxbrwwgv19NNPS6q+Z1tSUpLuvfdeTZo0yc/V4WSGYWj58uUaPny4v0tBHYqLi5WQkKCPPvpIl19+ub/L8So6NX5QUlKiuLg4f5cBBJyKigpt2LBBQ4YMcY7ZbDYNGTJEn3zyiR8rA4JXSUmJJLWIzx1CjY9t375ds2fP1rhx4/xdChBw9u/fr6qqKiUmJrqMJyYm6ocffvBTVUDwcjgcmjhxovr3769zzz3X3+V4HaGmiSZNmiTDMOr92rJli8tj9uzZo6uvvlo333yzxo4d66fKW4amvD8AEGomTJigr776Si+//LK/S/GJcH8XEKx++9vfKjMzs959unfv7vxzYWGhBg8erH79+un555/3cnVo7PuDwNChQweFhYVp3759LuP79u1Tx44d/VQVEJzuuecerVixQqtXr1aXLl38XY5PEGqaKD4+XvHx8Q3ad8+ePRo8eLDS0tI0f/582Ww0yLytMe8PAkdERITS0tL0wQcfOCeeOhwOffDBB7rnnnv8WxwQJEzT1L333qvly5dr1apVSklJ8XdJPkOo8bI9e/Zo0KBB6tatm2bOnKni4mLnNn7zDAy7du3Sf//7X+3atUtVVVXauHGjJOn0009XmzZt/FtcC/TAAw9o5MiR6tu3ry666CI9+eSTKisr06hRo/xdGk44fPiwtm/f7vx+x44d2rhxo+Li4tS1a1c/Vgap+pTTkiVL9MYbb6ht27bO+WixsbGKioryc3VeZsKr5s+fb0qy/EJgGDlypOX7k5ub6+/SWqzZs2ebXbt2NSMiIsyLLrrI/PTTT/1dEmrJzc21/DczcuRIf5cG06zzM2f+/Pn+Ls3ruE4NAAAICUzuAAAAIYFQAwAAQgKhBgAAhARCDQAACAmEGgAAEBIINQAAICQQagAAQEgg1AAAgJBAqAEQkD755BOFhYVp2LBh/i4FQJDgisIAAtKYMWPUpk0bzZs3T1u3blXnzp0t9zNNU1VVVQoP51Z2QEtHpwZAwDl8+LCWLl2q8ePHa9iwYcrOznZuW7VqlQzD0DvvvKO0tDTZ7XZ9/PHHcjgcmj59ulJSUhQVFaXzzjtPr7zyivNxVVVVGj16tHP7WWedpVmzZvnh1QHwFn61ARBw/vGPf+jss8/WWWedpREjRmjixImaPHmyDMNw7jNp0iTNnDlT3bt3V/v27TV9+nTl5ORo7ty5OuOMM7R69WqNGDFC8fHxGjhwoBwOh7p06aJly5bpf/7nf7R27Vrdfffd6tSpk2655RY/vloAnsLpJwABp3///rrlllt03333qbKyUp06ddKyZcs0aNAgrVq1SoMHD9brr7+u66+/XpJUXl6uuLg4vf/++7r00kudxxkzZoyOHDmiJUuWWD7PPffcox9++MGlowMgeNGpARBQtm7dqnXr1mn58uWSpPDwcN16662aN2+eBg0a5Nyvb9++zj9v375dR44c0ZVXXulyrIqKCp1//vnO75955hm9+OKL2rVrl44ePaqKigr16dPHq68HgO8QagAElHnz5qmystJlYrBpmrLb7Xr66aedY9HR0c4/Hz58WJL01ltv6bTTTnM5nt1ulyS9/PLLevDBB/XEE0/o0ksvVdu2bTVjxgx99tln3nw5AHyIUAMgYFRWVmrhwoV64okndNVVV7lsGz58uF566SWdffbZbo/r0aOH7Ha7du3apYEDB1oee82aNerXr59+/etfO8f+85//ePYFAPArQg2AgLFixQr9+OOPGj16tGJjY1223XjjjZo3b55mzJjh9ri2bdvqwQcf1P333y+Hw6EBAwaopKREa9asUUxMjEaOHKkzzjhDCxcu1MqVK5WSkqJFixZp/fr1SklJ8dXLA+BlLOkGEDDmzZunIUOGuAUaqTrU5OXl6YsvvrB87LRp0/TII49o+vTpOuecc3T11VfrrbfecoaWcePG6YYbbtCtt96qiy++WAcOHHDp2gAIfqx+AgAAIYFODQAACAmEGgAAEBIINQAAICQQagAAQEgg1AAAgJBAqAEAACGBUAMAAEICoQYAAIQEQg0AAAgJhBoAABASCDUAACAkEGoAAEBI+H+ASqn10QYkpQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_iters = 100\n",
    "for epoch in range(n_iters):\n",
    "    \n",
    "    # Forward pass\n",
    "    y_pred = model2(X)\n",
    "    \n",
    "    # Loss calculation\n",
    "    loss = criterion (y_pred, y)\n",
    "    \n",
    "    # Backward pass and updating the weights\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "\n",
    "    # Making the optimizer zero gradient before new step\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "    if (epoch+1) % 10 == 0:\n",
    "        print(f'epoch: {epoch+1}, loss = {loss.item():.4f}')\n",
    "\n",
    "# Plot\n",
    "predicted = model2(X).detach().numpy()\n",
    "\n",
    "plt.plot(X, y, 'g*')\n",
    "plt.plot(X, predicted, 'r')\n",
    "plt.title('Best Fit Line of Our Model')\n",
    "plt.xlabel('Area')\n",
    "plt.ylabel('Rent')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d0db3eb3",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}