1_Workflow.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMT+jlcNJaTKjVwAiY4e1ss",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/JaiSuryaPrabu/deep_learning/blob/main/1_Workflow.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Importing package"
      ],
      "metadata": {
        "id": "Jqvz_HOU8iAd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "from torch import nn\n",
        "from matplotlib import pyplot as plt\n",
        "\n",
        "torch.__version__"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "5tdiLAc98lZV",
        "outputId": "f9f85ac1-4534-4cbd-8bea-e79199fda7f9"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'2.0.1+cu118'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Data "
      ],
      "metadata": {
        "id": "oqs021gg8twp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# create a data\n",
        "weights = 0.7\n",
        "bias = 0.3\n",
        "\n",
        "# X is features and y is labels\n",
        "X = torch.arange(start = 0,end = 1, step = 0.01)\n",
        "y = weights * X + bias\n",
        "\n",
        "print(X.shape,y.shape)"
      ],
      "metadata": {
        "id": "VqsGE8ra8v9y",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "db15c00f-5ad1-4256-ce53-dccaa787dc33"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([100]) torch.Size([100])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# splitting the data\n",
        "training_percent = 0.75\n",
        "# 75 percent is training data and 25 percent is testing data\n",
        "num_of_training = int(training_percent * len(X))\n",
        "print(num_of_training)\n",
        "\n",
        "x_train,x_test = X[:num_of_training],X[num_of_training:]\n",
        "y_train,y_test = y[:num_of_training],y[num_of_training:]\n",
        "\n",
        "x_train.shape,x_test.shape,y_train.shape,y_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aQXwpzK8j0ip",
        "outputId": "6ed7829f-915b-4061-ae4c-5cc4b6f46c94"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "75\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(torch.Size([75]), torch.Size([25]), torch.Size([75]), torch.Size([25]))"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# ploting the data\n",
        "def plot_graph(Xtrain_data=x_train,\n",
        "               Xtest_data=x_test,\n",
        "               Ytrain_data=y_train,\n",
        "               Ytest_data=y_test,\n",
        "               predictions=None):\n",
        "    \n",
        "    plt.figure(figsize=(10,7))\n",
        "\n",
        "    # plotting training data\n",
        "    plt.scatter(Xtrain_data,Ytrain_data,c=\"b\",s=5,label=\"training data\")\n",
        "\n",
        "    # testing data\n",
        "    plt.scatter(Xtest_data,Ytest_data,c=\"r\",s=5,label=\"testing data\")\n",
        "\n",
        "    if predictions is not None:\n",
        "        plt.scatter(Xtest_data,predictions,c='g',s=5,label=\"predictions\")\n",
        "\n",
        "    plt.legend(prop={\"size\":10})\n",
        "\n",
        "plot_graph()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 599
        },
        "id": "Bg4efwOBlCuD",
        "outputId": "5b479c34-b9cf-4c92-e3db-88e4ee2d9001"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzoAAAJGCAYAAACTJvC6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABACUlEQVR4nO3dfXRV5Z0v8F8SSAIioEUSwCiKb1gRFCQD2lFngHR0cbW9c2Vsi0h96Qvaai5TQREqKrRX64WlTJ2xWrx2WrHWOl2jRSGVcVQqinKrt4j1BWEsBLE1USxEknP/SBNJJSEncHLePp+1zjqcnb3PeQ7uUr7s5/nugkQikQgAAIAcUpjuAQAAABxogg4AAJBzBB0AACDnCDoAAEDOEXQAAICcI+gAAAA5R9ABAAByTo90D6Azmpqa4ve//30cfPDBUVBQkO7hAAAAaZJIJOL999+PwYMHR2Fh+9dtsiLo/P73v4+Kiop0DwMAAMgQmzdvjsMPP7zdn2dF0Dn44IMjovnL9O3bN82jAQAA0qW+vj4qKipaM0J7siLotExX69u3r6ADAADsc0mLMgIAACDnCDoAAEDOEXQAAICckxVrdDqjqakpGhoa0j0MMljPnj2jqKgo3cMAAKAb5ETQaWhoiDfffDOamprSPRQyXP/+/aO8vNz9mAAAclzWB51EIhFbtmyJoqKiqKio6PCmQeSvRCIRH374YWzbti0iIgYNGpTmEQEAkEpZH3R2794dH374YQwePDh69+6d7uGQwXr16hUREdu2bYuBAweaxgYAkMOy/vJHY2NjREQUFxeneSRkg5Yw/NFHH6V5JAAApFLWB50W1lzQGc4TAID8kDNBBwAAoIWgkyOGDh0aixYt6vT+q1atioKCgnjvvfdSNqb2LF26NPr379/tnwsAQP4QdNLkrLPOiquuuuqAvd9zzz0Xl19+eaf3Hz9+fGzZsiX69et3wMaQSskGOQAA8lvWt67lskQiEY2NjdGjx77/Mx122GFJvXdxcXGUl5d3dWgAAJDRkr6i8+STT8bkyZNj8ODBUVBQEA8//PA+j1m1alWceuqpUVJSEsccc0wsXbq0C0PNHRdffHH8x3/8RyxevDgKCgqioKAgNm7c2Dqd7Je//GWMHj06SkpK4qmnnorXX389zjvvvCgrK4s+ffrEaaedFitXrmzznn95xaOgoCB+8IMfxOc+97no3bt3HHvssfGLX/yi9ed/OXWtZTrZY489FsOHD48+ffrEZz/72diyZUvrMbt3745vfOMb0b9///jUpz4V11xzTUybNi3OP//8Dr/v0qVL44gjjojevXvH5z73uXj33Xfb/Hxf3++ss86Kt956K66++urW36+IiHfffTcuvPDCGDJkSPTu3TtGjBgRP/nJT5L5TwEAQI5KOujs2LEjRo4cGUuWLOnU/m+++Wace+65cfbZZ8e6deviqquuiksvvTQee+yxpAebKxYvXhzjxo2Lyy67LLZs2RJbtmyJioqK1p/PmjUrvvOd78T69evj5JNPjg8++CDOOeecqKmpiRdffDE++9nPxuTJk2PTpk0dfs4NN9wQF1xwQfzmN7+Jc845J774xS/GH/7wh3b3//DDD+PWW2+N++67L5588snYtGlTzJw5s/Xn3/3ud+Nf//Vf44c//GE8/fTTUV9fv8+g++yzz8Yll1wSV1xxRaxbty7OPvvsuOmmm9rss6/v99BDD8Xhhx8e8+fPb/39iojYuXNnjB49Oh555JF4+eWX4/LLL4+pU6fGmjVrOhwTAAB5ILEfIiLx85//vMN9vvWtbyU+/elPt9k2ZcqURFVVVac/p66uLhERibq6uk/87E9/+lPit7/9beJPf/pTp99vbz76KJG44YZEYuLE5uePPtqvt9unM888M/HNb36zzbYnnngiERGJhx9+eJ/Hf/rTn07cfvvtra+PPPLIxP/+3/+79XVEJObMmdP6+oMPPkhEROKXv/xlm8/64x//mEgkEokf/vCHiYhIvPbaa63HLFmyJFFWVtb6uqysLHHLLbe0vt69e3fiiCOOSJx33nntjvPCCy9MnHPOOW22TZkyJdGvX7/9+n7tOffccxP/83/+z3Z/fqDOFwAA0qOjbLCnlJcRrF69OiZMmNBmW1VVVaxevbrdY3bt2hX19fVtHqm2YEHEt78dsWJF8/OCBSn/yHaNGTOmzesPPvggZs6cGcOHD4/+/ftHnz59Yv369fu8onPyySe3/vqggw6Kvn37xrZt29rdv3fv3jFs2LDW14MGDWrdv66uLmpra2Ps2LGtPy8qKorRo0d3OIb169dHZWVlm23jxo07IN+vsbExbrzxxhgxYkQceuih0adPn3jsscf2eRwAALkv5WUEW7dujbKysjbbysrKor6+Pv70pz9Fr169PnHMwoUL44Ybbkj10Np46qmIRKL514lE8+t0Oeigg9q8njlzZqxYsSJuvfXWOOaYY6JXr17x93//99HQ0NDh+/Ts2bPN64KCgmhqakpq/0TLb0oKdfX73XLLLbF48eJYtGhRjBgxIg466KC46qqr9nkcAAC5LyPrpWfPnh11dXWtj82bN6f8M884I+LPa9yjoKD5dSoVFxdHY2Njp/Z9+umn4+KLL47Pfe5zMWLEiCgvL4+NGzemdoB/oV+/flFWVhbPPfdc67bGxsZ44YUXOjxu+PDh8eyzz7bZ9utf/7rN6858v739fj399NNx3nnnxZe+9KUYOXJkHH300fHqq6924dsBAJBrUn5Fp7y8PGpra9tsq62tjb59++71ak5ERElJSZSUlKR6aG1ce23z81NPNYecltepMnTo0Hj22Wdj48aN0adPnzj00EPb3ffYY4+Nhx56KCZPnhwFBQVx/fXXd3hlJlWuvPLKWLhwYRxzzDFxwgknxO233x5//OMfW1vQ9uYb3/hGnH766XHrrbfGeeedF4899lgsX768zT6d+X5Dhw6NJ598Mv7hH/4hSkpKYsCAAXHsscfGgw8+GM8880wccsghcdttt0VtbW2ceOKJKfn+AABkj5Rf0Rk3blzU1NS02bZixYpPrNNItx49IubOjXj88ebnTty6Zr/MnDkzioqK4sQTT4zDDjusw3Ult912WxxyyCExfvz4mDx5clRVVcWpp56a2gHuxTXXXBMXXnhhXHTRRTFu3Ljo06dPVFVVRWlpabvH/NVf/VXcddddsXjx4hg5cmQ8/vjjMWfOnDb7dOb7zZ8/PzZu3BjDhg1rvWfQnDlz4tRTT42qqqo466yzory8fJ9V1wAAJGH37oj58yMmTWp+3r073SPqtIJEkoswPvjgg3jttdciIuKUU06J2267Lc4+++w49NBD44gjjojZs2fH22+/Hf/n//yfiGiulz7ppJNixowZ8eUvfzl+9atfxTe+8Y145JFHoqqqqlOfWV9fH/369Yu6urro27dvm5/t3Lkz3nzzzTjqqKM6/As3B15TU1MMHz48LrjggrjxxhvTPZxOcb4AACRh/vzmpq5Eonl9x7e/3XxVII06ygZ7Svq6xfPPPx9nn3126+vq6uqIiJg2bVosXbo0tmzZ0ubqxFFHHRWPPPJIXH311bF48eI4/PDD4wc/+EGnQw6Z46233orHH388zjzzzNi1a1fccccd8eabb8YXvvCFdA8NAIBUyKTGriQlHXTOOuusDpu4li5dutdjXnzxxWQ/igxTWFgYS5cujZkzZ0YikYiTTjopVq5cGcOHD0/30AAASIUzzohYufLjKzqpbuw6gFJeRkDuqKioiKeffjrdwwAAoLt0d2PXASToAABAPtu9O2LBgrZhpqWZq6WxKwsJOgAAkM8WLPi4cGDlyuZtWRpu9pSRNwwFAAC6SRYXDnRE0AEAgHx2xhnNRQMRWVc40BFT1wAAIJ9lceFARwQdAADIdTlaONARU9dy3MaNG6OgoCDWrVvX7Z+9atWqKCgoiPfee6/bPxsAgD20FA6sWNH8vGBBukeUcoJOmpx11llx1VVXHdD3vPjii+P8889vs62ioiK2bNkSJ5100gH9rFRJxe8LAEDey9HCgY4IOjmuqKgoysvLo0cPsxQBAPJWjhYOdETQSYOLL744/uM//iMWL14cBQUFUVBQEBs3boyIiJdffjn+7u/+Lvr06RNlZWUxderU2L59e+uxDz74YIwYMSJ69eoVn/rUp2LChAmxY8eO+Pa3vx333ntv/Nu//Vvre65ateoTU9dappPV1NTEmDFjonfv3jF+/PjYsGFDmzHedNNNMXDgwDj44IPj0ksvjVmzZsWoUaM6/F6PPvpoHHfccdGrV684++yzW79Ti3fffTcuvPDCGDJkSPTu3TtGjBgRP/nJT/b5+9LY2BiXXHJJHHXUUdGrV684/vjjY/HixV3+/QcAyDvXXts8ZW3ixObnHCkc6IigkwaLFy+OcePGxWWXXRZbtmyJLVu2REVFRbz33nvxN3/zN3HKKafE888/H8uXL4/a2tq44IILIiJiy5YtceGFF8aXv/zlWL9+faxatSo+//nPRyKRiJkzZ8YFF1wQn/3sZ1vfc/z48e2O4brrrovvfe978fzzz0ePHj3iy1/+cuvP/vVf/zVuvvnm+O53vxtr166NI444Ir7//e93+J02b94cn//852Py5Mmxbt261nC0p507d8bo0aPjkUceiZdffjkuv/zymDp1aqxZs6bD35empqY4/PDD46c//Wn89re/jblz58a1114bDzzwQFf/EwAA5J7duyPmz4+YNKn5effuj3/WUjjw+OPNz3kw2yf3v2FnddREcYD169cviouLo3fv3lFeXt66/Y477ohTTjklFuyxOOyee+6JioqKePXVV+ODDz6I3bt3x+c///k48sgjIyJixIgRrfv26tUrdu3a1eY923PzzTfHmWeeGRERs2bNinPPPTd27twZpaWlcfvtt8cll1wS06dPj4iIuXPnxuOPPx4ffPBBu+/3/e9/P4YNGxbf+973IiLi+OOPj5deeim++93vtu4zZMiQmDlzZuvrK6+8Mh577LF44IEHYuzYse3+vhQVFcUNN9zQ+vqoo46K1atXxwMPPNAaAgEA8l5L4UAiEbFyZfO2HGxT6yxBp0UGnBj/9//+33jiiSeiT58+n/jZ66+/HpMmTYq//du/jREjRkRVVVVMmjQp/v7v/z4OOeSQpD/r5JNPbv31oEGDIiJi27ZtccQRR8SGDRvi61//epv9x44dG7/61a/afb/169dHZWVlm23jxo1r87qxsTEWLFgQDzzwQLz99tvR0NAQu3btit69e+9zvEuWLIl77rknNm3aFH/605+ioaFhn1PpAADySh4WDnTE1LUWGXBifPDBB61Tv/Z8/O53v4u//uu/jqKiolixYkX88pe/jBNPPDFuv/32OP744+PNN99M+rN69uzZ+uuCPy9Ma2pqOmDfZW9uueWWWLx4cVxzzTXxxBNPxLp166KqqioaGho6PO7++++PmTNnxiWXXBKPP/54rFu3LqZPn77P4wAA8koeFg50xBWdFmec0XwlJ5HolhOjuLg4Ghsb22w79dRT42c/+1kMHTq03Za0goKCOP300+P000+PuXPnxpFHHhk///nPo7q6eq/v2RXHH398PPfcc3HRRRe1bnvuuec6PGb48OHxi1/8os22X//6121eP/3003HeeefFl770pYhoDlavvvpqnHjiia377O07PP300zF+/Pg2V5lef/315L4UAECuaykY2HMpRh5zRadFNzdRDB06NJ599tnYuHFjbN++PZqammLGjBnxhz/8IS688MJ47rnn4vXXX4/HHnsspk+fHo2NjfHss8/GggUL4vnnn49NmzbFQw89FO+8804MHz689T1/85vfxIYNG2L79u3x0UcfdWlsV155Zdx9991x7733xu9+97u46aab4je/+U3rlZ+9+epXvxq/+93v4h//8R9jw4YN8eMf/ziWLl3aZp9jjz02VqxYEc8880ysX78+vvKVr0Rtbe0+f1+OPfbYeP755+Oxxx6LV199Na6//vp9Bi8AgJykcKDTBJ0W3XxizJw5M4qKiuLEE0+Mww47LDZt2hSDBw+Op59+OhobG2PSpEkxYsSIuOqqq6J///5RWFgYffv2jSeffDLOOeecOO6442LOnDnxve99L/7u7/4uIiIuu+yyOP7442PMmDFx2GGHxdNPP92lsX3xi1+M2bNnx8yZM+PUU0+NN998My6++OIoLS1t95gjjjgifvazn8XDDz8cI0eOjDvvvLNNqUJExJw5c+LUU0+NqqqqOOuss6K8vPwTNzjd2+/LV77ylfj85z8fU6ZMicrKynj33Xc/sYYIACAvtKwrX7Gi+fkv/r7FxwoSiZaFKZmrvr4++vXrF3V1ddG3b982P9u5c2e8+eabcdRRR3X4F3H2z8SJE6O8vDzuu+++dA9lvzhfAICsNmlSc8hpMXFi8z/U55GOssGe8vt6Fnv14Ycfxp133hlVVVVRVFQUP/nJT2LlypWxYs//UQEA0P26eV15NhN0+ISCgoJ49NFH4+abb46dO3fG8ccfHz/72c9iwoQJ6R4aAEB+UzjQaYIOn9CrV69Y2XIvIQAAuldHN7JvWVfOPgk6AACQSTLgRva5IGda17KgU4EM4DwBADJeBtzIPhdkfdApKiqKiIiGhoY0j4Rs8OGHH0ZERM+ePdM8EgCAdpxxRnPRQITCgf2Q9VPXevToEb1794533nknevbsGYWFWZ/dSIFEIhEffvhhbNu2Lfr3798akAEA0qKjdTgKBw6IrA86BQUFMWjQoHjzzTfjrbfeSvdwyHD9+/eP8vLydA8DAMh3Ha3DUThwQGR90ImIKC4ujmOPPdb0NTrUs2dPV3IAgMxgHU7K5UTQiYgoLCx0p3sAALKDG3+mXM4EHQAAyBrW4aScoAMAAKngxp9pJegAAEAquPFnWuliBgCAVFA4kFaCDgAApIIbf6aVqWsAAJAKCgfSStABAICuUjiQsQQdAADoKoUDGcsaHQAA6CqFAxlL0AEAgK5SOJCxTF0DAICuUjiQsQQdAABoT0dlAxEKBzKYoAMAAO1RNpC1rNEBAID2KBvIWoIOAAC0R9lA1jJ1DQAA2qNsIGsJOgAA5LeOCgeUDWQtQQcAgPymcCAnWaMDAEB+UziQkwQdAADym8KBnGTqGgAA+U3hQE4SdAAAyH0KB/KOoAMAQO5TOJB3rNEBACD3KRzIO4IOAAC5T+FA3jF1DQCA3KdwIO8IOgAA5AaFA+xB0AEAIDcoHGAP1ugAAJAbFA6wB0EHAIDcoHCAPZi6BgBAblA4wB66dEVnyZIlMXTo0CgtLY3KyspYs2ZNu/t+9NFHMX/+/Bg2bFiUlpbGyJEjY/ny5V0eMAAAeWz37oj58yMmTWp+3r3745+1FA48/njzcw//pp/Pkg46y5Yti+rq6pg3b1688MILMXLkyKiqqopt27btdf85c+bEP//zP8ftt98ev/3tb+OrX/1qfO5zn4sXX3xxvwcPAECeaSkcWLGi+XnBgnSPiAxVkEi0rNjqnMrKyjjttNPijjvuiIiIpqamqKioiCuvvDJmzZr1if0HDx4c1113XcyYMaN123//7/89evXqFT/60Y869Zn19fXRr1+/qKuri759+yYzXAAAcsmkSc0hp8XEic1XcMgbnc0GSV3RaWhoiLVr18aECRM+foPCwpgwYUKsXr16r8fs2rUrSktL22zr1atXPNVBC8auXbuivr6+zQMAABQO0FlJTVzcvn17NDY2RllZWZvtZWVl8corr+z1mKqqqrjtttvir//6r2PYsGFRU1MTDz30UDQ2Nrb7OQsXLowbbrghmaEBAJAPFA7QSSmvl168eHEce+yxccIJJ0RxcXFcccUVMX369CgsbP+jZ8+eHXV1da2PzZs3p3qYAABkCoUDHABJnRkDBgyIoqKiqK2tbbO9trY2ysvL93rMYYcdFg8//HDs3Lkz3n333Rg8eHDMmjUrjj766HY/p6SkJEpKSpIZGgAAuaKlcCCRiFi5snnb3LlpHRLZJ6krOsXFxTF69Oioqalp3dbU1BQ1NTUxbty4Do8tLS2NIUOGxO7du+NnP/tZnHfeeV0bMQAAue2pp5pDTkTzcwdru6E9SU9dq66ujrvuuivuvffeWL9+fXzta1+LHTt2xPTp0yMi4qKLLorZs2e37v/ss8/GQw89FG+88Ub853/+Z3z2s5+Npqam+Na3vnXgvgUAALlD4QAHQNKTGqdMmRLvvPNOzJ07N7Zu3RqjRo2K5cuXtxYUbNq0qc36m507d8acOXPijTfeiD59+sQ555wT9913X/Tv3/+AfQkAAHKIwgEOgKTvo5MO7qMDAJBjdu9uXouzZ5hRLEAndDYbOJsAAOh+CgdIsZTXSwMAwCcoHCDFBB0AALqfwgFSzNQ1AAC6n8IBUkzQAQAgNToqHOjRw5ocUkrQAQAgNRQOkEbW6AAAkBoKB0gjQQcAgNRQOEAamboGAEBqKBwgjQQdAAC6TuEAGUrQAQCg6xQOkKGs0QEAoOsUDpChBB0AALpO4QAZytQ1AAC6TuEAGUrQAQCgfR2VDUQoHCBjCToAALRP2QBZyhodAADap2yALCXoAADQPmUDZClT1wAAaJ+yAbKUoAMAkO86KhxQNkCWEnQAAPKdwgFykDU6AAD5TuEAOUjQAQDIdwoHyEGmrgEA5DuFA+QgQQcAIB8oHCDPCDoAAPlA4QB5xhodAIB8oHCAPCPoAADkA4UD5BlT1wAA8oHCAfKMoAMAkCsUDkArQQcAIFcoHIBW1ugAAOQKhQPQStABAMgVCgeglalrAAC5QuEAtBJ0AACyicIB6BRBBwAgmygcgE6xRgcAIJsoHIBOEXQAALKJwgHoFFPXAACyicIB6BRBBwAg0ygcgP0m6AAAZBqFA7DfrNEBAMg0Cgdgvwk6AACZRuEA7DdT1wAAMo3CAdhvgg4AQDooHICUEnQAANJB4QCklDU6AADpoHAAUkrQAQBIB4UDkFKmrgEApIPCAUgpQQcAIEU66htQOACpJegAAKSIvgFIH2t0AABSRN8ApI+gAwCQIvoGIH1MXQMA6KIO1+CEvgFIJ0EHAKCL9rUGR98ApI+pawAAXWQNDmQuQQcAoIuswYHMZeoaAEAXWYMDmUvQAQDoQEeFA9bgQObq0tS1JUuWxNChQ6O0tDQqKytjzZo1He6/aNGiOP7446NXr15RUVERV199dezcubNLAwYA6E4thQMrVjQ/L1iQ7hEBnZF00Fm2bFlUV1fHvHnz4oUXXoiRI0dGVVVVbNu2ba/7//jHP45Zs2bFvHnzYv369XH33XfHsmXL4lrXdgGALKBwALJT0kHntttui8suuyymT58eJ554Ytx5553Ru3fvuOeee/a6/zPPPBOnn356fOELX4ihQ4fGpEmT4sILL9znVSAAgEygcACyU1JrdBoaGmLt2rUxe/bs1m2FhYUxYcKEWL169V6PGT9+fPzoRz+KNWvWxNixY+ONN96IRx99NKZOndru5+zatSt27drV+rq+vj6ZYQIAHDAKByA7JRV0tm/fHo2NjVFWVtZme1lZWbzyyit7PeYLX/hCbN++Pc4444xIJBKxe/fu+OpXv9rh1LWFCxfGDTfckMzQAAC6TOEA5J6U30dn1apVsWDBgvinf/qneOGFF+Khhx6KRx55JG688cZ2j5k9e3bU1dW1PjZv3pzqYQIAeUzhAOSepK7oDBgwIIqKiqK2trbN9tra2igvL9/rMddff31MnTo1Lr300oiIGDFiROzYsSMuv/zyuO6666Kw8JNZq6SkJEpKSpIZGgBAlykcgNyT1BWd4uLiGD16dNTU1LRua2pqipqamhg3btxej/nwww8/EWaKiooiIiLR8icKAEAaKRyA3JP0DUOrq6tj2rRpMWbMmBg7dmwsWrQoduzYEdOnT4+IiIsuuiiGDBkSCxcujIiIyZMnx2233RannHJKVFZWxmuvvRbXX399TJ48uTXwAACkk8IByD1JB50pU6bEO++8E3Pnzo2tW7fGqFGjYvny5a0FBZs2bWpzBWfOnDlRUFAQc+bMibfffjsOO+ywmDx5ctx8880H7lsAAOyDwgHILwWJLJg/Vl9fH/369Yu6urro27dvuocDAGSh+fObiwYSiebpad/+tnAD2aiz2SDlrWsAAJlA4QDkF0EHAMgLCgcgvyS9RgcAIBspHID8IugAADlD4QDQQtABAHLGggUfFw6sXNm8TbiB/GSNDgCQMxQOAC0EHQAgZygcAFqYugYA5AyFA0ALQQcAyCoKB4DOEHQAgKyicADoDGt0AICsonAA6AxBBwDIKgoHgM4wdQ0AyCoKB4DOEHQAgIyjcADYX4IOAJBxFA4A+8saHQAg4ygcAPaXoAMAZByFA8D+MnUNAMg4CgeA/SXoAABpoXAASCVBBwBIC4UDQCpZowMApIXCASCVBB0AIC0UDgCpZOoaAJAWCgeAVBJ0AICUUTgApIugAwCkjMIBIF2s0QEAUkbhAJAugg4AkDIKB4B0MXUNAEgZhQNAugg6AECXdVQ2EKFwAEgfQQcA6DJlA0CmskYHAOgyZQNAphJ0AIAuUzYAZCpT1wCALlM2AGQqQQcA6FBHhQPKBoBMJegAAB1SOABkI2t0AIAOKRwAspGgAwB0SOEAkI1MXQMAOqRwAMhGgg4AoHAAyDmCDgCgcADIOdboAAAKB4CcI+gAAAoHgJxj6hoAoHAAyDmCDgDkCYUDQD4RdAAgTygcAPKJNToAkCcUDgD5RNABgDyhcADIJ6auAUCeUDgA5BNBBwByiMIBgGaCDgDkEIUDAM2s0QGAHKJwAKCZoAMAOUThAEAzU9cAIIcoHABoJugAQJZROACwb4IOAGQZhQMA+2aNDgBkGYUDAPsm6ABAllE4ALBvpq4BQJZROACwb4IOAGQghQMA+0fQAYAMpHAAYP90aY3OkiVLYujQoVFaWhqVlZWxZs2advc966yzoqCg4BOPc889t8uDBoBcp3AAYP8kHXSWLVsW1dXVMW/evHjhhRdi5MiRUVVVFdu2bdvr/g899FBs2bKl9fHyyy9HUVFR/I//8T/2e/AAkKsUDgDsn4JEouXfizqnsrIyTjvttLjjjjsiIqKpqSkqKiriyiuvjFmzZu3z+EWLFsXcuXNjy5YtcdBBB3XqM+vr66Nfv35RV1cXffv2TWa4AJCVOlqjA5DPOpsNkvojs6GhIdauXRuzZ89u3VZYWBgTJkyI1atXd+o97r777viHf/iHDkPOrl27YteuXa2v6+vrkxkmAGQFhQMAqZNU0Nm+fXs0NjZGWVlZm+1lZWXxyiuv7PP4NWvWxMsvvxx33313h/stXLgwbrjhhmSGBgBZR+EAQOp06w1D77777hgxYkSMHTu2w/1mz54ddXV1rY/Nmzd30wgBoPsoHABInaSCzoABA6KoqChqa2vbbK+trY3y8vIOj92xY0fcf//9cckll+zzc0pKSqJv375tHgCQaxQOAKROUlPXiouLY/To0VFTUxPnn39+RDSXEdTU1MQVV1zR4bE//elPY9euXfGlL32py4MFgFxy7bXNz3uu0QHgwEi6v6W6ujqmTZsWY8aMibFjx8aiRYtix44dMX369IiIuOiii2LIkCGxcOHCNsfdfffdcf7558enPvWpAzNyAMhw+2pOUzgAkDpJB50pU6bEO++8E3Pnzo2tW7fGqFGjYvny5a0FBZs2bYrCwrYz4jZs2BBPPfVUPP744wdm1ACQBZQNAKRP0vfRSQf30QEgG02aFLFixcevJ06M8G9+APuns9mgW1vXACCfKBsASB/3WAaAFFE2AJA+gg4A7IeOCgeUDQCkj6ADAPtB4QBAZrJGBwD2w1NPNYeciObnp55K73gAaCboAMB+UDgAkJlMXQOA/aBwACAzCToAsA8KBwCyj6ADAPugcAAg+1ijAwD7oHAAIPsIOgCwDwoHALKPqWsAEB2vw1E4AJB9BB0AiI7X4SgcAMg+pq4BQFiHA5BrBB0ACOtwAHKNqWsAENbhAOQaQQeAvOHGnwD5Q9ABIG+48SdA/rBGB4C8oXAAIH8IOgDkDYUDAPnD1DUA8obCAYD8IegAkFMUDgAQIegAkGMUDgAQYY0OADlG4QAAEYIOADlG4QAAEaauAZBjFA4AECHoAJCFFA4AsC+CDgBZR+EAAPtijQ4AWUfhAAD7IugAkHUUDgCwL6auAZB1FA4AsC+CDgAZSeEAAPtD0AEgIykcAGB/WKMDQEZSOADA/hB0AMhICgcA2B+mrgGQkRQOALA/BB0A0kbhAACpIugAkDYKBwBIFWt0AEgbhQMApIqgA0DaKBwAIFVMXQMgbRQOAJAqgg4AKdNR2UCEwgEAUkfQASBllA0AkC7W6ACQMsoGAEgXQQeAlFE2AEC6mLoGQMooGwAgXQQdAPZLR4UDygYASBdBB4D9onAAgExkjQ4A+0XhAACZSNABYL8oHAAgE5m6BsB+UTgAQCYSdADYJ4UDAGQbQQeAfVI4AEC2sUYHgH1SOABAthF0ANgnhQMAZBtT1wDYJ4UDAGQbQQeAiFA4AEBuEXQAiAiFAwDkFmt0AIgIhQMA5BZBB4CIUDgAQG7pUtBZsmRJDB06NEpLS6OysjLWrFnT4f7vvfdezJgxIwYNGhQlJSVx3HHHxaOPPtqlAQOQGtde2zx1beLE5meFAwBks6TX6Cxbtiyqq6vjzjvvjMrKyli0aFFUVVXFhg0bYuDAgZ/Yv6GhISZOnBgDBw6MBx98MIYMGRJvvfVW9O/f/0CMH4AkKBwAIF8UJBItM7I7p7KyMk477bS44447IiKiqakpKioq4sorr4xZs2Z9Yv8777wzbrnllnjllVeiZ8+eXRpkfX199OvXL+rq6qJv375deg8AIubP/7hwoKCg+dfCDQDZpLPZIKmpaw0NDbF27dqYMGHCx29QWBgTJkyI1atX7/WYX/ziFzFu3LiYMWNGlJWVxUknnRQLFiyIxsbGdj9n165dUV9f3+YBwP5TOABAvkgq6Gzfvj0aGxujrKyszfaysrLYunXrXo9544034sEHH4zGxsZ49NFH4/rrr4/vfe97cdNNN7X7OQsXLox+/fq1PioqKpIZJgDtUDgAQL5I+X10mpqaYuDAgfEv//IvUVRUFKNHj4633347brnllpg3b95ej5k9e3ZUV1e3vq6vrxd2AA6AloKBPdfoAEAuSiroDBgwIIqKiqK2trbN9tra2igvL9/rMYMGDYqePXtGUVFR67bhw4fH1q1bo6GhIYqLiz9xTElJSZSUlCQzNAD+TOEAACQ5da24uDhGjx4dNTU1rduampqipqYmxo0bt9djTj/99Hjttdeiqampddurr74agwYN2mvIAWD/LFjQXDKwYkXz84IF6R4RAHS/pO+jU11dHXfddVfce++9sX79+vja174WO3bsiOnTp0dExEUXXRSzZ89u3f9rX/ta/OEPf4hvfvOb8eqrr8YjjzwSCxYsiBkzZhy4bwFAK4UDANCFNTpTpkyJd955J+bOnRtbt26NUaNGxfLly1sLCjZt2hSFhR/np4qKinjsscfi6quvjpNPPjmGDBkS3/zmN+Oaa645cN8CgFZnnBGxcuXHFdIKBwDIR0nfRycd3EcHoPM6WqMDANmus9nA//UBZCGFAwDQMUEHIAu1FA4kEs3T1CKEGwDYU9JlBACkn8IBAOiYoAOQhc44o7loIELhAADsjalrAFno2mubn/dcowMAfEzQAchQCgcAoOsEHYAMpXAAALrOGh2ADKVwAAC6TtAByFAKBwCg60xdA8hQCgcAoOsEHYA06ahsIELhAADsD0EHIE2UDQBA6lijA5AmygYAIHUEHYA0UTYAAKlj6hpAmigbAIDUEXQAUqijwgFlAwCQOoIOQAopHACA9LBGByCFFA4AQHoIOgAppHAAANLD1DWAFFI4AADpIegA7CeFAwCQeQQdgP2kcAAAMo81OgD7SeEAAGQeQQdgPykcAIDMY+oawH5SOAAAmUfQAegEhQMAkF0EHYBOUDgAANnFGh2ATlA4AADZRdAB6ASFAwCQXUxdA+gEhQMAkF0EHYA/UzgAALlD0AH4M4UDAJA7rNEB+DOFAwCQOwQdgD9TOAAAucPUNSCvdLQOR+EAAOQOQQfIKx2tw1E4AAC5w9Q1IK9YhwMA+UHQAfKKdTgAkB9MXQPyinU4AJAfBB0g57jxJwAg6AA5x40/AQBrdICco3AAABB0gJyjcAAAMHUNyDkKBwAAQQfISgoHAICOCDpAVlI4AAB0xBodICspHAAAOiLoAFlJ4QAA0BFT14CspHAAAOiIoANkLIUDAEBXCTpAxlI4AAB0lTU6QMZSOAAAdJWgA2QshQMAQFeZugZkLIUDAEBXCTpA2nRUNhChcAAA6DpBB0gbZQMAQKpYowOkjbIBACBVBB0gbZQNAACpYuoakDbKBgCAVBF0gJTqqHBA2QAAkCpdmrq2ZMmSGDp0aJSWlkZlZWWsWbOm3X2XLl0aBQUFbR6lpaVdHjCQXVoKB1asaH5esCDdIwIA8kHSQWfZsmVRXV0d8+bNixdeeCFGjhwZVVVVsW3btnaP6du3b2zZsqX18dZbb+3XoIHsoXAAAEiHpIPObbfdFpdddllMnz49TjzxxLjzzjujd+/ecc8997R7TEFBQZSXl7c+ysrKOvyMXbt2RX19fZsHkJ0UDgAA6ZBU0GloaIi1a9fGhAkTPn6DwsKYMGFCrF69ut3jPvjggzjyyCOjoqIizjvvvPh//+//dfg5CxcujH79+rU+KioqkhkmkEGuvbZ5ytrEic3PCgcAgO6QVNDZvn17NDY2fuKKTFlZWWzdunWvxxx//PFxzz33xL/927/Fj370o2hqaorx48fHf/3Xf7X7ObNnz466urrWx+bNm5MZJtDNdu+OmD8/YtKk5ufduz/+WUvhwOOPNz/3UIECAHSDlP+VY9y4cTFu3LjW1+PHj4/hw4fHP//zP8eNN96412NKSkqipKQk1UMDDpCWwoFEImLlyuZt2tQAgHRK6orOgAEDoqioKGpra9tsr62tjfLy8k69R8+ePeOUU06J1157LZmPBjKYwgEAINMkFXSKi4tj9OjRUVNT07qtqakpampq2ly16UhjY2O89NJLMWjQoORGCmQshQMAQKZJeupadXV1TJs2LcaMGRNjx46NRYsWxY4dO2L69OkREXHRRRfFkCFDYuHChRERMX/+/Pirv/qrOOaYY+K9996LW265Jd5666249NJLD+w3AdKmpWBgz5uCAgCkU9JBZ8qUKfHOO+/E3LlzY+vWrTFq1KhYvnx5a0HBpk2borDw4wtFf/zjH+Oyyy6LrVu3xiGHHBKjR4+OZ555Jk488cQD9y2AlNu9u3ktzp5hpqVYoKVwAAAgUxQkEi0z6zNXfX199OvXL+rq6qJv377pHg7kpfnzPy4cKCho/rVwAwB0t85mg6RvGArkJ4UDAEA2EXSATlE4AABkE7fuAzpF4QAAkE0EHaCVwgEAIFcIOkCrBQs+LhxYubJ5m3ADAGQja3SAVgoHAIBcIegArRQOAAC5wtQ1oJXCAQAgVwg6kGcUDgAA+UDQgTyjcAAAyAfW6ECeUTgAAOQDQQfyjMIBACAfmLoGeUbhAACQDwQdyEEKBwCAfCfoQA5SOAAA5DtrdCAHKRwAAPKdoAM5SOEAAJDvTF2DHKRwAADId4IOZCmFAwAA7RN0IEspHAAAaJ81OpClFA4AALRP0IEspXAAAKB9pq5BllI4AADQPkEHMpjCAQCArhF0IIMpHAAA6BprdCCDKRwAAOgaQQcymMIBAICuMXUNMpjCAQCArhF0II06KhuIUDgAANBVgg6kkbIBAIDUsEYH0kjZAABAagg6kEbKBgAAUsPUNUgjZQMAAKkh6ECKdVQ4oGwAACA1BB1IMYUDAADdzxodSDGFAwAA3U/QgRRTOAAA0P1MXYMUUzgAAND9BB04ABQOAABkFkEHDgCFAwAAmcUaHTgAFA4AAGQWQQcOAIUDAACZxdQ1OAAUDgAAZBZBBzpJ4QAAQPYQdKCTFA4AAGQPa3SgkxQOAABkD0EHOknhAABA9jB1DTpJ4QAAQPYQdGAPCgcAAHKDoAN7UDgAAJAbrNGBPSgcAADIDYIO7EHhAABAbjB1DfagcAAAIDcIOuQdhQMAALlP0CHvKBwAAMh91uiQdxQOAADkPkGHvKNwAAAg95m6Rt5ROAAAkPsEHXKSwgEAgPzWpalrS5YsiaFDh0ZpaWlUVlbGmjVrOnXc/fffHwUFBXH++ed35WOh01oKB1asaH5esCDdIwIAoDslHXSWLVsW1dXVMW/evHjhhRdi5MiRUVVVFdu2bevwuI0bN8bMmTPjM5/5TJcHC52lcAAAIL8lHXRuu+22uOyyy2L69Olx4oknxp133hm9e/eOe+65p91jGhsb44tf/GLccMMNcfTRR+/XgKEzFA4AAOS3pNboNDQ0xNq1a2P27Nmt2woLC2PChAmxevXqdo+bP39+DBw4MC655JL4z//8z31+zq5du2LXrl2tr+vr65MZJnmio3U4CgcAAPJbUkFn+/bt0djYGGVlZW22l5WVxSuvvLLXY5566qm4++67Y926dZ3+nIULF8YNN9yQzNDIQx3d+FPhAABAfkvpfXTef//9mDp1atx1110xYMCATh83e/bsqKura31s3rw5haMkW1mHAwBAe5K6ojNgwIAoKiqK2traNttra2ujvLz8E/u//vrrsXHjxpg8eXLrtqampuYP7tEjNmzYEMOGDfvEcSUlJVFSUpLM0MhDZ5zRfCUnkbAOBwCAtpIKOsXFxTF69OioqalprYhuamqKmpqauOKKKz6x/wknnBAvvfRSm21z5syJ999/PxYvXhwVFRVdHzl5zzocAADak/QNQ6urq2PatGkxZsyYGDt2bCxatCh27NgR06dPj4iIiy66KIYMGRILFy6M0tLSOOmkk9oc379//4iIT2yHv9RR2UCEdTgAALQv6aAzZcqUeOedd2Lu3LmxdevWGDVqVCxfvry1oGDTpk1RWJjSpT/kiY7KBgAAoCMFiUTLcu7MVV9fH/369Yu6urro27dvuodDN5k0KWLFio9fT5wY8fjj6RsPAADp19ls4NILGctNPwEA6Kqkp65Bd1E2AABAVwk6pFVHhQPKBgAA6CpBh7RSOAAAQCpYo0NaPfVUc8iJaH5+6qn0jgcAgNwg6JBWCgcAAEgFU9dIK4UDAACkgqBDyikcAACguwk6pJzCAQAAups1OqScwgEAALqboEPKKRwAAKC7mbpGyikcAACguwk6HBAKBwAAyCSCDgeEwgEAADKJNTocEAoHAADIJIIOB4TCAQAAMompaxwQCgcAAMgkgg6dpnAAAIBsIejQaQoHAADIFtbo0GkKBwAAyBaCDp2mcAAAgGxh6hqdpnAAAIBsIejQhsIBAABygaBDGwoHAADIBdbo0IbCAQAAcoGgQxsKBwAAyAWmrtGGwgEAAHKBoJOHFA4AAJDrBJ08pHAAAIBcZ41OHlI4AABArhN08pDCAQAAcp2pa3lI4QAAALlO0MlRCgcAAMhngk6OUjgAAEA+s0YnRykcAAAgnwk6OUrhAAAA+czUtRylcAAAgHwm6GQxhQMAALB3gk4WUzgAAAB7Z41OFlM4AAAAeyfoZDGFAwAAsHemrmUxhQMAALB3gk4G66hsIELhAAAAtEfQyWDKBgAAoGus0clgygYAAKBrBJ0MpmwAAAC6xtS1DKZsAAAAukbQSbOOCgeUDQAAQNcIOmmmcAAAAA48a3TSTOEAAAAceIJOmikcAACAA8/UtTRTOAAAAAeeoNMNFA4AAED3EnS6gcIBAADoXtbodAOFAwAA0L0EnW6gcAAAALqXqWvdQOEAAAB0L0HnAFE4AAAAmUPQOUAUDgAAQOawRucAUTgAAACZo0tBZ8mSJTF06NAoLS2NysrKWLNmTbv7PvTQQzFmzJjo379/HHTQQTFq1Ki47777ujzgTKVwAAAAMkfSU9eWLVsW1dXVceedd0ZlZWUsWrQoqqqqYsOGDTFw4MBP7H/ooYfGddddFyeccEIUFxfHv//7v8f06dNj4MCBUVVVdUC+RCZQOAAAAJmjIJFomXDVOZWVlXHaaafFHXfcERERTU1NUVFREVdeeWXMmjWrU+9x6qmnxrnnnhs33nhjp/avr6+Pfv36RV1dXfTt2zeZ4R5QHRUOAAAAqdfZbJDUX9MbGhpi7dq1MXv27NZthYWFMWHChFi9evU+j08kEvGrX/0qNmzYEN/97nfb3W/Xrl2xa9eu1tf19fXJDDNlFA4AAEB2SGqNzvbt26OxsTHKysrabC8rK4utW7e2e1xdXV306dMniouL49xzz43bb789Jk6c2O7+CxcujH79+rU+KioqkhlmyigcAACA7NAtrWsHH3xwrFu3Lp577rm4+eabo7q6OlatWtXu/rNnz466urrWx+bNm7tjmPukcAAAALJDUlPXBgwYEEVFRVFbW9tme21tbZSXl7d7XGFhYRxzzDERETFq1KhYv359LFy4MM4666y97l9SUhIlJSXJDK1bKBwAAIDskFTQKS4ujtGjR0dNTU2cf/75EdFcRlBTUxNXXHFFp9+nqampzRqcbNGjhzU5AACQDZLuDKuuro5p06bFmDFjYuzYsbFo0aLYsWNHTJ8+PSIiLrroohgyZEgsXLgwIprX24wZMyaGDRsWu3btikcffTTuu++++P73v39gvwkAAMCfJR10pkyZEu+8807MnTs3tm7dGqNGjYrly5e3FhRs2rQpCgs/XvqzY8eO+PrXvx7/9V//Fb169YoTTjghfvSjH8WUKVMO3LcAAADYQ9L30UmHTLmPDgAAkF6dzQbd0roGAADQnQQdAAAg5wg6AABAzhF0AACAnCPoAAAAOUfQAQAAco6gAwAA5BxBBwAAyDmCDgAAkHMEHQAAIOcIOgAAQM4RdAAAgJwj6AAAADlH0AEAAHKOoAMAAOQcQQcAAMg5PdI9gM5IJBIREVFfX5/mkQAAAOnUkglaMkJ7siLovP/++xERUVFRkeaRAAAAmeD999+Pfv36tfvzgsS+olAGaGpqit///vdx8MEHR0FBQVrHUl9fHxUVFbF58+bo27dvWsdC9nDe0FXOHbrCeUNXOG/oqu4+dxKJRLz//vsxePDgKCxsfyVOVlzRKSwsjMMPPzzdw2ijb9++/hAgac4busq5Q1c4b+gK5w1d1Z3nTkdXclooIwAAAHKOoAMAAOQcQSdJJSUlMW/evCgpKUn3UMgizhu6yrlDVzhv6ArnDV2VqedOVpQRAAAAJMMVHQAAIOcIOgAAQM4RdAAAgJwj6AAAADlH0AEAAHKOoLMXS5YsiaFDh0ZpaWlUVlbGmjVrOtz/pz/9aZxwwglRWloaI0aMiEcffbSbRkomSea8ueuuu+Izn/lMHHLIIXHIIYfEhAkT9nmekbuS/TOnxf333x8FBQVx/vnnp3aAZKRkz5v33nsvZsyYEYMGDYqSkpI47rjj/P9VHkr2vFm0aFEcf/zx0atXr6ioqIirr746du7c2U2jJRM8+eSTMXny5Bg8eHAUFBTEww8/vM9jVq1aFaeeemqUlJTEMcccE0uXLk35OPdG0PkLy5Yti+rq6pg3b1688MILMXLkyKiqqopt27btdf9nnnkmLrzwwrjkkkvixRdfjPPPPz/OP//8ePnll7t55KRTsufNqlWr4sILL4wnnngiVq9eHRUVFTFp0qR4++23u3nkpFuy506LjRs3xsyZM+Mzn/lMN42UTJLsedPQ0BATJ06MjRs3xoMPPhgbNmyIu+66K4YMGdLNIyedkj1vfvzjH8esWbNi3rx5sX79+rj77rtj2bJlce2113bzyEmnHTt2xMiRI2PJkiWd2v/NN9+Mc889N84+++xYt25dXHXVVXHppZfGY489luKR7kWCNsaOHZuYMWNG6+vGxsbE4MGDEwsXLtzr/hdccEHi3HPPbbOtsrIy8ZWvfCWl4ySzJHve/KXdu3cnDj744MS9996bqiGSobpy7uzevTsxfvz4xA9+8IPEtGnTEuedd143jJRMkux58/3vfz9x9NFHJxoaGrpriGSgZM+bGTNmJP7mb/6mzbbq6urE6aefntJxkrkiIvHzn/+8w32+9a1vJT796U+32TZlypREVVVVCke2d67o7KGhoSHWrl0bEyZMaN1WWFgYEyZMiNWrV+/1mNWrV7fZPyKiqqqq3f3JPV05b/7Shx9+GB999FEceuihqRomGair5878+fNj4MCBcckll3THMMkwXTlvfvGLX8S4ceNixowZUVZWFieddFIsWLAgGhsbu2vYpFlXzpvx48fH2rVrW6e3vfHGG/Hoo4/GOeec0y1jJjtl0t+Ne3T7J2aw7du3R2NjY5SVlbXZXlZWFq+88spej9m6dete99+6dWvKxklm6cp585euueaaGDx48Cf+YCC3deXceeqpp+Luu++OdevWdcMIyURdOW/eeOON+NWvfhVf/OIX49FHH43XXnstvv71r8dHH30U8+bN645hk2ZdOW++8IUvxPbt2+OMM86IRCIRu3fvjq9+9aumrtGh9v5uXF9fH3/605+iV69e3TYWV3Qgzb7zne/E/fffHz//+c+jtLQ03cMhg73//vsxderUuOuuu2LAgAHpHg5ZpKmpKQYOHBj/8i//EqNHj44pU6bEddddF3feeWe6h0YGW7VqVSxYsCD+6Z/+KV544YV46KGH4pFHHokbb7wx3UODTnFFZw8DBgyIoqKiqK2tbbO9trY2ysvL93pMeXl5UvuTe7py3rS49dZb4zvf+U6sXLkyTj755FQOkwyU7Lnz+uuvx8aNG2Py5Mmt25qamiIiokePHrFhw4YYNmxYagdN2nXlz5xBgwZFz549o6ioqHXb8OHDY+vWrdHQ0BDFxcUpHTPp15Xz5vrrr4+pU6fGpZdeGhERI0aMiB07dsTll18e1113XRQW+vdyPqm9vxv37du3W6/mRLii00ZxcXGMHj06ampqWrc1NTVFTU1NjBs3bq/HjBs3rs3+ERErVqxod39yT1fOm4iI//W//lfceOONsXz58hgzZkx3DJUMk+y5c8IJJ8RLL70U69ata338t//231qbbSoqKrpz+KRJV/7MOf300+O1115rDcYREa+++moMGjRIyMkTXTlvPvzww0+EmZawnEgkUjdYslpG/d242+sPMtz999+fKCkpSSxdujTx29/+NnH55Zcn+vfvn9i6dWsikUgkpk6dmpg1a1br/k8//XSiR48eiVtvvTWxfv36xLx58xI9e/ZMvPTSS+n6CqRBsufNd77znURxcXHiwQcfTGzZsqX18f7776frK5AmyZ47f0nrWn5K9rzZtGlT4uCDD05cccUViQ0bNiT+/d//PTFw4MDETTfdlK6vQBoke97MmzcvcfDBByd+8pOfJN54443E448/nhg2bFjiggsuSNdXIA3ef//9xIsvvph48cUXExGRuO222xIvvvhi4q233kokEonErFmzElOnTm3d/4033kj07t078Y//+I+J9evXJ5YsWZIoKipKLF++vNvHLujsxe2335444ogjEsXFxYmxY8cmfv3rX7f+7Mwzz0xMmzatzf4PPPBA4rjjjksUFxcnPv3pTyceeeSRbh4xmSCZ8+bII49MRMQnHvPmzev+gZN2yf6ZsydBJ38le94888wzicrKykRJSUni6KOPTtx8882J3bt3d/OoSbdkzpuPPvoo8e1vfzsxbNiwRGlpaaKioiLx9a9/PfHHP/6x+wdO2jzxxBN7/TtLy7kybdq0xJlnnvmJY0aNGpUoLi5OHH300Ykf/vCH3T7uRCKRKEgkXHsEAAByizU6AABAzhF0AACAnCPoAAAAOUfQAQAAco6gAwAA5BxBBwAAyDmCDgAAkHMEHQAAIOcIOgAAQM4RdAAAgJwj6AAAADnn/wP5N7M5LRHFiQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Build model\n"
      ],
      "metadata": {
        "id": "3kvoIFYvnD65"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# building a linear regression in pytorch\n",
        "\n",
        "class LinearRegression(nn.Module):\n",
        "\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "\n",
        "        self.weights = nn.Parameter(torch.randn(1,dtype=torch.float,requires_grad=True))\n",
        "        self.bias = nn.Parameter(torch.randn(1,dtype=torch.float,requires_grad=True))\n",
        "\n",
        "    def forward(self,input):\n",
        "        output = self.weights * input + self.bias\n",
        "        return output"
      ],
      "metadata": {
        "id": "4RoNU8B3nLw9"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "torch.manual_seed(42)\n",
        "model = LinearRegression()\n",
        "\n",
        "# to see the values of parameter such as weighs and bias\n",
        "print(model.weights.data)\n",
        "print(model.bias.data)\n",
        "\n",
        "print(model.parameters())\n",
        "print(model.state_dict())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8tOfCYhtoRmk",
        "outputId": "73b99477-421e-4603-d829-8173a2e42109"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([0.3367])\n",
            "tensor([0.1288])\n",
            "<generator object Module.parameters at 0x7f2a6c2d7df0>\n",
            "OrderedDict([('weights', tensor([0.3367])), ('bias', tensor([0.1288]))])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# try to use parameters() and the state_dict()\n",
        "\n",
        "# predicting the test data without optimization\n",
        "\n",
        "with torch.inference_mode():\n",
        "    y_preds = model(x_test)\n",
        "\n",
        "    plot_graph(predictions=y_preds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 599
        },
        "id": "ui0FIzyIpiqV",
        "outputId": "954b14a8-37a8-4bd8-b3a9-c954cd3e59b2"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Train model"
      ],
      "metadata": {
        "id": "kCtiZ-GxqyCw"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "* Loss functions \n",
        "    * Regression - Mean Absolute Error - `torch.nn.L1Loss()`\n",
        "    * Classification - Binary Cross Entropy - `torch.nn.BCELoss()`\n",
        "* Optimizer\n",
        "    * Stochastic gradient descent - `torch.optim.SGD()`\n",
        "    * Adam - `torch.optim.Adam()`"
      ],
      "metadata": {
        "id": "Y85kiTDLrHZ-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# creating a loss function and optimizer\n",
        "\n",
        "loss = torch.nn.L1Loss()\n",
        "\n",
        "optimizer = torch.optim.SGD(params=model.parameters(),lr=0.01)"
      ],
      "metadata": {
        "id": "B3LNDGGBqzSp"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Training Loop"
      ],
      "metadata": {
        "id": "Q2ZNz36Nt20k"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "1. Forward pass\n",
        "    * `model(x_train)`\n",
        "2. Calculate the loss\n",
        "    * `loss = loss(y_pred,y_train)`\n",
        "3. Zero gradients\n",
        "    * Each training steps needs 0 gradient values so it needs to be refreshed\n",
        "    *   `optimizer.zero_grad()`\n",
        "4. Perform backpropagation on loss\n",
        "    * `loss.backward()`\n",
        "5. Update the optimizer \n",
        "    * `optimizer.step()`"
      ],
      "metadata": {
        "id": "mGBIJzvbt6ip"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Testing loop"
      ],
      "metadata": {
        "id": "wR7swNGMvIaE"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "1. Forward pass\n",
        "2. Calculate loss\n",
        "3. Calculate evaluation metrics (optional)"
      ],
      "metadata": {
        "id": "F-6mcnUtww-a"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# training the model\n",
        "\n",
        "epochs = 150\n",
        "\n",
        "epoch_count = []\n",
        "train_loss_values = []\n",
        "test_loss_values = []\n",
        "\n",
        "for epoch in range(epochs):\n",
        "\n",
        "    # training phase\n",
        "\n",
        "    # training mode\n",
        "    model.train()\n",
        "\n",
        "    # 1. Forward pass\n",
        "    predicted = model(x_train)\n",
        "\n",
        "    # 2. Calculate the loss\n",
        "    loss_value = loss(predicted,y_train)\n",
        "\n",
        "    # 3. Zero gradients\n",
        "    optimizer.zero_grad()\n",
        "\n",
        "    # 4. loss backpropagation\n",
        "    loss_value.backward()\n",
        "\n",
        "    # 5. Updata optimizer\n",
        "    optimizer.step()\n",
        "\n",
        "    # testing phase at every 1/10 of training\n",
        "\n",
        "    if epoch % 10 == 0:\n",
        "        with torch.inference_mode():\n",
        "            test_pred = model(x_test)\n",
        "            test_loss = loss(test_pred,y_test.type(torch.float))\n",
        "            epoch_count.append(epoch)\n",
        "            train_loss_values.append(loss_value.detach().numpy())\n",
        "            test_loss_values.append(test_loss.detach().numpy())\n",
        "            print(f\"Epoch {epoch} Training Loss {loss_value} Testing loss {test_loss}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_yPNpXdKt57J",
        "outputId": "0214cec8-face-4054-9b0a-59d4c08e3af6"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 0 Training Loss 0.2942461669445038 Testing loss 0.46083199977874756\n",
            "Epoch 10 Training Loss 0.18055614829063416 Testing loss 0.3286420404911041\n",
            "Epoch 20 Training Loss 0.07795001566410065 Testing loss 0.20529867708683014\n",
            "Epoch 30 Training Loss 0.04955681413412094 Testing loss 0.14155007898807526\n",
            "Epoch 40 Training Loss 0.04368527978658676 Testing loss 0.11515693366527557\n",
            "Epoch 50 Training Loss 0.04042576253414154 Testing loss 0.10273885726928711\n",
            "Epoch 60 Training Loss 0.03734460100531578 Testing loss 0.09459733217954636\n",
            "Epoch 70 Training Loss 0.03427761048078537 Testing loss 0.08645584434270859\n",
            "Epoch 80 Training Loss 0.03120969794690609 Testing loss 0.07867151498794556\n",
            "Epoch 90 Training Loss 0.02814178168773651 Testing loss 0.07088716328144073\n",
            "Epoch 100 Training Loss 0.025073867291212082 Testing loss 0.0631028339266777\n",
            "Epoch 110 Training Loss 0.022005952894687653 Testing loss 0.05531850829720497\n",
            "Epoch 120 Training Loss 0.018938036635518074 Testing loss 0.04753417149186134\n",
            "Epoch 130 Training Loss 0.015870125964283943 Testing loss 0.039749838411808014\n",
            "Epoch 140 Training Loss 0.012802702374756336 Testing loss 0.03160834312438965\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Plotting the loss"
      ],
      "metadata": {
        "id": "tEKnNNyXylaJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(epoch_count, train_loss_values, label=\"Train loss\")\n",
        "plt.plot(epoch_count, test_loss_values, label=\"Test loss\")\n",
        "plt.title(\"Training and test loss curves\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "9QqrodtKyn5E",
        "outputId": "3fd07c91-6710-42cf-e74e-7102ac629824"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f2a6cc0dea0>"
            ]
          },
          "metadata": {},
          "execution_count": 30
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Predcitions of trained model"
      ],
      "metadata": {
        "id": "bRb55fyRyy4O"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "with torch.inference_mode():\n",
        "    y_preds = model(x_test)\n",
        "\n",
        "    plot_graph(predictions=y_preds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 599
        },
        "id": "LoGiMoZ8y1YS",
        "outputId": "91f2625f-a98c-43c7-c8b0-2f33388eab02"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Saving and loading"
      ],
      "metadata": {
        "id": "mrUjTKprzGqq"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The three methods are :    \n",
        "* `torch.save`\n",
        "* `torch.load`\n",
        "* `torch.nn.Module.load_state_dict`"
      ],
      "metadata": {
        "id": "FcZk6JqUzTvi"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Saving"
      ],
      "metadata": {
        "id": "A7DUlTPSzJ23"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# saving the state_dict of the model\n",
        "model_name = \"Model.pth\"\n",
        "torch.save(obj = model.state_dict(),f=model_name)"
      ],
      "metadata": {
        "id": "kW2rRXcQzK1-"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Loading"
      ],
      "metadata": {
        "id": "x8XuNihK1s_c"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# create a new model\n",
        "new_model = LinearRegression()\n",
        "\n",
        "new_model.load_state_dict(torch.load(f=model_name))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LmGsi0NZ1uSk",
        "outputId": "f0cbdd1d-6de1-4738-a807-b8568434d254"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# checking the two models parameters\n",
        "model.state_dict(),new_model.state_dict()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oEdc-5WR2Oja",
        "outputId": "010666a4-d2c6-48fe-95e2-5708bcd3738e"
      },
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(OrderedDict([('weights', tensor([0.6483])), ('bias', tensor([0.3204]))]),\n",
              " OrderedDict([('weights', tensor([0.6483])), ('bias', tensor([0.3204]))]))"
            ]
          },
          "metadata": {},
          "execution_count": 36
        }
      ]
    }
  ]
}