NeuralNetwork.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "NeuralNetwork.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPVtmDk7jNFe7oi/VmvItdX"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "e3Jk1l5RjCuy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aBZdjm7rJqxD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def softmax(a):\n",
        "  e_pa = np.exp(a) #Vector\n",
        "  ans = e_pa/np.sum(e_pa,axis=1,keepdims=True)\n",
        "  return ans"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2U85HgTU0JFt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def loss(y_oht,p):\n",
        "  return -np.mean(y_oht*np.log(p))\n",
        "\n",
        "def one_hot(y,depth):\n",
        "  #depth here stands for the number of labels\n",
        "  m=y.shape[0] #number of examples\n",
        "  y_oht=np.zeros((m,depth))\n",
        "  y_oht[np.arange(m),y]=1 #puts 1 at the given class in a row, while the other vaues remain 0\n",
        "  return y_oht"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vm0R44sMwFGK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class NeuralNetwork:\n",
        "\n",
        "  def __init__(self,input_size,layer,output_size):\n",
        "    model= {} #Empty Dictionary\n",
        "    np.random.seed(0)\n",
        "    model['W1']=np.random.randn(input_size,layer[0])\n",
        "    model['b1']=np.zeros((1,layer[0]))\n",
        "\n",
        "    model['W2']=np.random.randn(layer[0],layer[1])\n",
        "    model['b2']=np.zeros((1,layer[1]))\n",
        "\n",
        "    model['W3']=np.random.randn(layer[1],output_size)\n",
        "    model['b3']=np.zeros((1,output_size))\n",
        "\n",
        "    self.model=model\n",
        "    self.activation_outputs= None\n",
        "  \n",
        "  def forward(self,X): #forward propagation implementation\n",
        "    \n",
        "    W1,W2,W3=self.model[\"W1\"],self.model[\"W2\"],self.model[\"W3\"]\n",
        "    b1,b2,b3=self.model[\"b1\"],self.model[\"b2\"],self.model[\"b3\"]\n",
        "    #print(W1.shape,b1.shape)# W1=4,3 b1=3,1\n",
        "    Z1=np.dot(X,W1)+b1\n",
        "    #print(Z1.shape) 2x3\n",
        "    a1=np.tanh(Z1)\n",
        "    #print(a1.shape)2x3\n",
        "    #print(Z1)\n",
        "    Z2=np.dot(a1,W2)+b2\n",
        "    a2=np.tanh(Z2)\n",
        "\n",
        "    Z3=np.dot(a2,W3)+b3\n",
        "    #a3=np.tanh(Z3)\n",
        "    y_=softmax(Z3)\n",
        "    self.activation_outputs=(a1,a2,y_) #the activation values obtained from each layer\n",
        "    return y_\n",
        "  \n",
        "  def backward(self,x,y,learning_rate):\n",
        "    \n",
        "    W1,W2,W3 = self.model['W1'],self.model['W2'],self.model['W3']\n",
        "    b1, b2, b3 = self.model['b1'],self.model['b2'],self.model['b3']\n",
        "    m=x.shape[0] #number of training examples\n",
        "    a1,a2,y_= self.activation_outputs\n",
        "    #starting with the last layer\n",
        "    delta3=y_-y\n",
        "    dw3 = np.dot(a2.T,delta3)\n",
        "    db3 = np.sum(delta3,axis=0)\n",
        "\n",
        "    delta2 = (1-np.square(a2))*np.dot(delta3,W3.T)\n",
        "    dw2 = np.dot(a1.T,delta2)\n",
        "    db2 = np.sum(delta2,axis=0)\n",
        "\n",
        "    delta1 = (1-np.square(a1))*np.dot(delta2,W2.T)\n",
        "    dw1 = np.dot(X.T,delta1)\n",
        "    db1 = np.sum(delta1,axis=0)\n",
        "\n",
        "    self.model[\"W1\"]  -= learning_rate*dw1\n",
        "    self.model['b1']  -= learning_rate*db1\n",
        "        \n",
        "    self.model[\"W2\"]  -= learning_rate*dw2\n",
        "    self.model['b2']  -= learning_rate*db2\n",
        "        \n",
        "    self.model[\"W3\"]  -= learning_rate*dw3\n",
        "    self.model['b3']  -= learning_rate*db3\n",
        "  \n",
        "  def predict(self,x):\n",
        "    y_out=self.forward(x) #returns a row vector in 1 hot form to represent the output class\n",
        "    return np.argmax(y_out,axis=1) \n",
        " \n",
        "  def summary(self):\n",
        "    \n",
        "    W1,W2,W3 = self.model['W1'],self.model['W2'],self.model['W3']\n",
        "    a1,a2,y_ = self.activation_outputs\n",
        "        \n",
        "    print(\"W1 \",W1.shape)\n",
        "    print(\"A1 \",a1.shape)\n",
        "        \n",
        "    print(\"W2 \",W2.shape)\n",
        "    print(\"A2 \",a2.shape)\n",
        "        \n",
        "    print(\"W3 \",W3.shape)\n",
        "    print(\"Y_ \",y_.shape)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Djg3EaiPINWs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.datasets import make_circles\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "96AhYHILuKPU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X,Y= make_circles(n_samples=500,shuffle=True,random_state=1,noise=0.2,factor=0.2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X5f8fv6VubeM",
        "colab_type": "code",
        "outputId": "bf700134-cbee-4721-d4a2-36a3e5a3cba3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "plt.scatter(X[:,0],X[:,1],c=Y)\n",
        "plt.show()"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD4CAYAAADhNOGaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyOb8H7ogEPH52nbnAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EakpjonoJ7GY",
        "colab_type": "text"
      },
      "source": [
        "# Training the Network\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GLQkV8THKLmZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model= NeuralNetwork(input_size=2,layer=[10,5],output_size=2) #2 classes in output, 2 features in input"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gM53puHekBwF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(X,Y,model,max_epochs=500,learning_rate=0.001,logs=True):\n",
        "  #logs if set will print the loss as the training takes place\n",
        "  training_loss=[]\n",
        "  classes=2\n",
        "  Y_OHT= one_hot(Y,classes) #converting the output vector into one hot form\n",
        "  #print(\"One hot shape=\",Y_OHT.shape)\n",
        "\n",
        "  for ix in range(max_epochs):\n",
        "    Y_=model.forward(X) #current prediction\n",
        "    l=loss(Y_OHT,Y_)\n",
        "    training_loss.append(l)\n",
        "    model.backward(X,Y_OHT,learning_rate)\n",
        "\n",
        "    if(logs):\n",
        "      print(\"Epoch %d Loss %.4f\"%(ix,l))\n",
        "  return training_loss"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3bRkH_tKohyc",
        "colab_type": "code",
        "outputId": "6e347c2d-6017-4cec-8605-85eac6e2d94a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "losses=train(X,Y,model)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 0 Loss 0.3571\n",
            "Epoch 1 Loss 0.3554\n",
            "Epoch 2 Loss 0.2593\n",
            "Epoch 3 Loss 0.2407\n",
            "Epoch 4 Loss 0.2258\n",
            "Epoch 5 Loss 0.2132\n",
            "Epoch 6 Loss 0.2020\n",
            "Epoch 7 Loss 0.1919\n",
            "Epoch 8 Loss 0.1827\n",
            "Epoch 9 Loss 0.1742\n",
            "Epoch 10 Loss 0.1664\n",
            "Epoch 11 Loss 0.1593\n",
            "Epoch 12 Loss 0.1527\n",
            "Epoch 13 Loss 0.1467\n",
            "Epoch 14 Loss 0.1411\n",
            "Epoch 15 Loss 0.1360\n",
            "Epoch 16 Loss 0.1313\n",
            "Epoch 17 Loss 0.1270\n",
            "Epoch 18 Loss 0.1230\n",
            "Epoch 19 Loss 0.1193\n",
            "Epoch 20 Loss 0.1159\n",
            "Epoch 21 Loss 0.1127\n",
            "Epoch 22 Loss 0.1098\n",
            "Epoch 23 Loss 0.1070\n",
            "Epoch 24 Loss 0.1045\n",
            "Epoch 25 Loss 0.1021\n",
            "Epoch 26 Loss 0.0999\n",
            "Epoch 27 Loss 0.0978\n",
            "Epoch 28 Loss 0.0958\n",
            "Epoch 29 Loss 0.0940\n",
            "Epoch 30 Loss 0.0922\n",
            "Epoch 31 Loss 0.0906\n",
            "Epoch 32 Loss 0.0891\n",
            "Epoch 33 Loss 0.0876\n",
            "Epoch 34 Loss 0.0862\n",
            "Epoch 35 Loss 0.0849\n",
            "Epoch 36 Loss 0.0837\n",
            "Epoch 37 Loss 0.0825\n",
            "Epoch 38 Loss 0.0814\n",
            "Epoch 39 Loss 0.0803\n",
            "Epoch 40 Loss 0.0793\n",
            "Epoch 41 Loss 0.0783\n",
            "Epoch 42 Loss 0.0774\n",
            "Epoch 43 Loss 0.0765\n",
            "Epoch 44 Loss 0.0756\n",
            "Epoch 45 Loss 0.0748\n",
            "Epoch 46 Loss 0.0740\n",
            "Epoch 47 Loss 0.0732\n",
            "Epoch 48 Loss 0.0725\n",
            "Epoch 49 Loss 0.0718\n",
            "Epoch 50 Loss 0.0711\n",
            "Epoch 51 Loss 0.0705\n",
            "Epoch 52 Loss 0.0699\n",
            "Epoch 53 Loss 0.0693\n",
            "Epoch 54 Loss 0.0687\n",
            "Epoch 55 Loss 0.0681\n",
            "Epoch 56 Loss 0.0676\n",
            "Epoch 57 Loss 0.0671\n",
            "Epoch 58 Loss 0.0666\n",
            "Epoch 59 Loss 0.0661\n",
            "Epoch 60 Loss 0.0656\n",
            "Epoch 61 Loss 0.0651\n",
            "Epoch 62 Loss 0.0647\n",
            "Epoch 63 Loss 0.0643\n",
            "Epoch 64 Loss 0.0638\n",
            "Epoch 65 Loss 0.0634\n",
            "Epoch 66 Loss 0.0630\n",
            "Epoch 67 Loss 0.0627\n",
            "Epoch 68 Loss 0.0623\n",
            "Epoch 69 Loss 0.0619\n",
            "Epoch 70 Loss 0.0616\n",
            "Epoch 71 Loss 0.0612\n",
            "Epoch 72 Loss 0.0609\n",
            "Epoch 73 Loss 0.0606\n",
            "Epoch 74 Loss 0.0602\n",
            "Epoch 75 Loss 0.0599\n",
            "Epoch 76 Loss 0.0596\n",
            "Epoch 77 Loss 0.0593\n",
            "Epoch 78 Loss 0.0591\n",
            "Epoch 79 Loss 0.0588\n",
            "Epoch 80 Loss 0.0585\n",
            "Epoch 81 Loss 0.0582\n",
            "Epoch 82 Loss 0.0580\n",
            "Epoch 83 Loss 0.0577\n",
            "Epoch 84 Loss 0.0575\n",
            "Epoch 85 Loss 0.0572\n",
            "Epoch 86 Loss 0.0570\n",
            "Epoch 87 Loss 0.0568\n",
            "Epoch 88 Loss 0.0565\n",
            "Epoch 89 Loss 0.0563\n",
            "Epoch 90 Loss 0.0561\n",
            "Epoch 91 Loss 0.0559\n",
            "Epoch 92 Loss 0.0557\n",
            "Epoch 93 Loss 0.0555\n",
            "Epoch 94 Loss 0.0553\n",
            "Epoch 95 Loss 0.0551\n",
            "Epoch 96 Loss 0.0549\n",
            "Epoch 97 Loss 0.0547\n",
            "Epoch 98 Loss 0.0545\n",
            "Epoch 99 Loss 0.0544\n",
            "Epoch 100 Loss 0.0542\n",
            "Epoch 101 Loss 0.0540\n",
            "Epoch 102 Loss 0.0538\n",
            "Epoch 103 Loss 0.0537\n",
            "Epoch 104 Loss 0.0535\n",
            "Epoch 105 Loss 0.0534\n",
            "Epoch 106 Loss 0.0532\n",
            "Epoch 107 Loss 0.0530\n",
            "Epoch 108 Loss 0.0529\n",
            "Epoch 109 Loss 0.0527\n",
            "Epoch 110 Loss 0.0526\n",
            "Epoch 111 Loss 0.0525\n",
            "Epoch 112 Loss 0.0523\n",
            "Epoch 113 Loss 0.0522\n",
            "Epoch 114 Loss 0.0520\n",
            "Epoch 115 Loss 0.0519\n",
            "Epoch 116 Loss 0.0518\n",
            "Epoch 117 Loss 0.0517\n",
            "Epoch 118 Loss 0.0515\n",
            "Epoch 119 Loss 0.0514\n",
            "Epoch 120 Loss 0.0513\n",
            "Epoch 121 Loss 0.0512\n",
            "Epoch 122 Loss 0.0510\n",
            "Epoch 123 Loss 0.0509\n",
            "Epoch 124 Loss 0.0508\n",
            "Epoch 125 Loss 0.0507\n",
            "Epoch 126 Loss 0.0506\n",
            "Epoch 127 Loss 0.0505\n",
            "Epoch 128 Loss 0.0504\n",
            "Epoch 129 Loss 0.0503\n",
            "Epoch 130 Loss 0.0502\n",
            "Epoch 131 Loss 0.0501\n",
            "Epoch 132 Loss 0.0500\n",
            "Epoch 133 Loss 0.0499\n",
            "Epoch 134 Loss 0.0498\n",
            "Epoch 135 Loss 0.0497\n",
            "Epoch 136 Loss 0.0496\n",
            "Epoch 137 Loss 0.0495\n",
            "Epoch 138 Loss 0.0494\n",
            "Epoch 139 Loss 0.0493\n",
            "Epoch 140 Loss 0.0492\n",
            "Epoch 141 Loss 0.0491\n",
            "Epoch 142 Loss 0.0490\n",
            "Epoch 143 Loss 0.0489\n",
            "Epoch 144 Loss 0.0489\n",
            "Epoch 145 Loss 0.0488\n",
            "Epoch 146 Loss 0.0487\n",
            "Epoch 147 Loss 0.0486\n",
            "Epoch 148 Loss 0.0485\n",
            "Epoch 149 Loss 0.0484\n",
            "Epoch 150 Loss 0.0484\n",
            "Epoch 151 Loss 0.0483\n",
            "Epoch 152 Loss 0.0482\n",
            "Epoch 153 Loss 0.0481\n",
            "Epoch 154 Loss 0.0481\n",
            "Epoch 155 Loss 0.0480\n",
            "Epoch 156 Loss 0.0479\n",
            "Epoch 157 Loss 0.0478\n",
            "Epoch 158 Loss 0.0478\n",
            "Epoch 159 Loss 0.0477\n",
            "Epoch 160 Loss 0.0476\n",
            "Epoch 161 Loss 0.0476\n",
            "Epoch 162 Loss 0.0475\n",
            "Epoch 163 Loss 0.0474\n",
            "Epoch 164 Loss 0.0474\n",
            "Epoch 165 Loss 0.0473\n",
            "Epoch 166 Loss 0.0472\n",
            "Epoch 167 Loss 0.0472\n",
            "Epoch 168 Loss 0.0471\n",
            "Epoch 169 Loss 0.0471\n",
            "Epoch 170 Loss 0.0470\n",
            "Epoch 171 Loss 0.0469\n",
            "Epoch 172 Loss 0.0469\n",
            "Epoch 173 Loss 0.0468\n",
            "Epoch 174 Loss 0.0468\n",
            "Epoch 175 Loss 0.0467\n",
            "Epoch 176 Loss 0.0466\n",
            "Epoch 177 Loss 0.0466\n",
            "Epoch 178 Loss 0.0465\n",
            "Epoch 179 Loss 0.0465\n",
            "Epoch 180 Loss 0.0464\n",
            "Epoch 181 Loss 0.0464\n",
            "Epoch 182 Loss 0.0463\n",
            "Epoch 183 Loss 0.0463\n",
            "Epoch 184 Loss 0.0462\n",
            "Epoch 185 Loss 0.0462\n",
            "Epoch 186 Loss 0.0461\n",
            "Epoch 187 Loss 0.0461\n",
            "Epoch 188 Loss 0.0460\n",
            "Epoch 189 Loss 0.0460\n",
            "Epoch 190 Loss 0.0459\n",
            "Epoch 191 Loss 0.0459\n",
            "Epoch 192 Loss 0.0458\n",
            "Epoch 193 Loss 0.0458\n",
            "Epoch 194 Loss 0.0457\n",
            "Epoch 195 Loss 0.0457\n",
            "Epoch 196 Loss 0.0456\n",
            "Epoch 197 Loss 0.0456\n",
            "Epoch 198 Loss 0.0455\n",
            "Epoch 199 Loss 0.0455\n",
            "Epoch 200 Loss 0.0454\n",
            "Epoch 201 Loss 0.0454\n",
            "Epoch 202 Loss 0.0453\n",
            "Epoch 203 Loss 0.0453\n",
            "Epoch 204 Loss 0.0453\n",
            "Epoch 205 Loss 0.0452\n",
            "Epoch 206 Loss 0.0452\n",
            "Epoch 207 Loss 0.0451\n",
            "Epoch 208 Loss 0.0451\n",
            "Epoch 209 Loss 0.0450\n",
            "Epoch 210 Loss 0.0450\n",
            "Epoch 211 Loss 0.0450\n",
            "Epoch 212 Loss 0.0449\n",
            "Epoch 213 Loss 0.0449\n",
            "Epoch 214 Loss 0.0448\n",
            "Epoch 215 Loss 0.0448\n",
            "Epoch 216 Loss 0.0448\n",
            "Epoch 217 Loss 0.0447\n",
            "Epoch 218 Loss 0.0447\n",
            "Epoch 219 Loss 0.0447\n",
            "Epoch 220 Loss 0.0446\n",
            "Epoch 221 Loss 0.0446\n",
            "Epoch 222 Loss 0.0445\n",
            "Epoch 223 Loss 0.0445\n",
            "Epoch 224 Loss 0.0445\n",
            "Epoch 225 Loss 0.0444\n",
            "Epoch 226 Loss 0.0444\n",
            "Epoch 227 Loss 0.0444\n",
            "Epoch 228 Loss 0.0443\n",
            "Epoch 229 Loss 0.0443\n",
            "Epoch 230 Loss 0.0443\n",
            "Epoch 231 Loss 0.0442\n",
            "Epoch 232 Loss 0.0442\n",
            "Epoch 233 Loss 0.0442\n",
            "Epoch 234 Loss 0.0441\n",
            "Epoch 235 Loss 0.0441\n",
            "Epoch 236 Loss 0.0441\n",
            "Epoch 237 Loss 0.0440\n",
            "Epoch 238 Loss 0.0440\n",
            "Epoch 239 Loss 0.0440\n",
            "Epoch 240 Loss 0.0439\n",
            "Epoch 241 Loss 0.0439\n",
            "Epoch 242 Loss 0.0439\n",
            "Epoch 243 Loss 0.0438\n",
            "Epoch 244 Loss 0.0438\n",
            "Epoch 245 Loss 0.0438\n",
            "Epoch 246 Loss 0.0437\n",
            "Epoch 247 Loss 0.0437\n",
            "Epoch 248 Loss 0.0437\n",
            "Epoch 249 Loss 0.0436\n",
            "Epoch 250 Loss 0.0436\n",
            "Epoch 251 Loss 0.0436\n",
            "Epoch 252 Loss 0.0436\n",
            "Epoch 253 Loss 0.0435\n",
            "Epoch 254 Loss 0.0435\n",
            "Epoch 255 Loss 0.0435\n",
            "Epoch 256 Loss 0.0434\n",
            "Epoch 257 Loss 0.0434\n",
            "Epoch 258 Loss 0.0434\n",
            "Epoch 259 Loss 0.0434\n",
            "Epoch 260 Loss 0.0433\n",
            "Epoch 261 Loss 0.0433\n",
            "Epoch 262 Loss 0.0433\n",
            "Epoch 263 Loss 0.0432\n",
            "Epoch 264 Loss 0.0432\n",
            "Epoch 265 Loss 0.0432\n",
            "Epoch 266 Loss 0.0432\n",
            "Epoch 267 Loss 0.0431\n",
            "Epoch 268 Loss 0.0431\n",
            "Epoch 269 Loss 0.0431\n",
            "Epoch 270 Loss 0.0431\n",
            "Epoch 271 Loss 0.0430\n",
            "Epoch 272 Loss 0.0430\n",
            "Epoch 273 Loss 0.0430\n",
            "Epoch 274 Loss 0.0429\n",
            "Epoch 275 Loss 0.0429\n",
            "Epoch 276 Loss 0.0429\n",
            "Epoch 277 Loss 0.0429\n",
            "Epoch 278 Loss 0.0428\n",
            "Epoch 279 Loss 0.0428\n",
            "Epoch 280 Loss 0.0428\n",
            "Epoch 281 Loss 0.0428\n",
            "Epoch 282 Loss 0.0427\n",
            "Epoch 283 Loss 0.0427\n",
            "Epoch 284 Loss 0.0427\n",
            "Epoch 285 Loss 0.0427\n",
            "Epoch 286 Loss 0.0426\n",
            "Epoch 287 Loss 0.0426\n",
            "Epoch 288 Loss 0.0426\n",
            "Epoch 289 Loss 0.0426\n",
            "Epoch 290 Loss 0.0426\n",
            "Epoch 291 Loss 0.0425\n",
            "Epoch 292 Loss 0.0425\n",
            "Epoch 293 Loss 0.0425\n",
            "Epoch 294 Loss 0.0425\n",
            "Epoch 295 Loss 0.0424\n",
            "Epoch 296 Loss 0.0424\n",
            "Epoch 297 Loss 0.0424\n",
            "Epoch 298 Loss 0.0424\n",
            "Epoch 299 Loss 0.0423\n",
            "Epoch 300 Loss 0.0423\n",
            "Epoch 301 Loss 0.0423\n",
            "Epoch 302 Loss 0.0423\n",
            "Epoch 303 Loss 0.0423\n",
            "Epoch 304 Loss 0.0422\n",
            "Epoch 305 Loss 0.0422\n",
            "Epoch 306 Loss 0.0422\n",
            "Epoch 307 Loss 0.0422\n",
            "Epoch 308 Loss 0.0422\n",
            "Epoch 309 Loss 0.0421\n",
            "Epoch 310 Loss 0.0421\n",
            "Epoch 311 Loss 0.0421\n",
            "Epoch 312 Loss 0.0421\n",
            "Epoch 313 Loss 0.0420\n",
            "Epoch 314 Loss 0.0420\n",
            "Epoch 315 Loss 0.0420\n",
            "Epoch 316 Loss 0.0420\n",
            "Epoch 317 Loss 0.0420\n",
            "Epoch 318 Loss 0.0419\n",
            "Epoch 319 Loss 0.0419\n",
            "Epoch 320 Loss 0.0419\n",
            "Epoch 321 Loss 0.0419\n",
            "Epoch 322 Loss 0.0419\n",
            "Epoch 323 Loss 0.0418\n",
            "Epoch 324 Loss 0.0418\n",
            "Epoch 325 Loss 0.0418\n",
            "Epoch 326 Loss 0.0418\n",
            "Epoch 327 Loss 0.0418\n",
            "Epoch 328 Loss 0.0417\n",
            "Epoch 329 Loss 0.0417\n",
            "Epoch 330 Loss 0.0417\n",
            "Epoch 331 Loss 0.0417\n",
            "Epoch 332 Loss 0.0417\n",
            "Epoch 333 Loss 0.0417\n",
            "Epoch 334 Loss 0.0416\n",
            "Epoch 335 Loss 0.0416\n",
            "Epoch 336 Loss 0.0416\n",
            "Epoch 337 Loss 0.0416\n",
            "Epoch 338 Loss 0.0416\n",
            "Epoch 339 Loss 0.0415\n",
            "Epoch 340 Loss 0.0415\n",
            "Epoch 341 Loss 0.0415\n",
            "Epoch 342 Loss 0.0415\n",
            "Epoch 343 Loss 0.0415\n",
            "Epoch 344 Loss 0.0415\n",
            "Epoch 345 Loss 0.0414\n",
            "Epoch 346 Loss 0.0414\n",
            "Epoch 347 Loss 0.0414\n",
            "Epoch 348 Loss 0.0414\n",
            "Epoch 349 Loss 0.0414\n",
            "Epoch 350 Loss 0.0413\n",
            "Epoch 351 Loss 0.0413\n",
            "Epoch 352 Loss 0.0413\n",
            "Epoch 353 Loss 0.0413\n",
            "Epoch 354 Loss 0.0413\n",
            "Epoch 355 Loss 0.0413\n",
            "Epoch 356 Loss 0.0412\n",
            "Epoch 357 Loss 0.0412\n",
            "Epoch 358 Loss 0.0412\n",
            "Epoch 359 Loss 0.0412\n",
            "Epoch 360 Loss 0.0412\n",
            "Epoch 361 Loss 0.0412\n",
            "Epoch 362 Loss 0.0411\n",
            "Epoch 363 Loss 0.0411\n",
            "Epoch 364 Loss 0.0411\n",
            "Epoch 365 Loss 0.0411\n",
            "Epoch 366 Loss 0.0411\n",
            "Epoch 367 Loss 0.0411\n",
            "Epoch 368 Loss 0.0410\n",
            "Epoch 369 Loss 0.0410\n",
            "Epoch 370 Loss 0.0410\n",
            "Epoch 371 Loss 0.0410\n",
            "Epoch 372 Loss 0.0410\n",
            "Epoch 373 Loss 0.0410\n",
            "Epoch 374 Loss 0.0410\n",
            "Epoch 375 Loss 0.0409\n",
            "Epoch 376 Loss 0.0409\n",
            "Epoch 377 Loss 0.0409\n",
            "Epoch 378 Loss 0.0409\n",
            "Epoch 379 Loss 0.0409\n",
            "Epoch 380 Loss 0.0409\n",
            "Epoch 381 Loss 0.0408\n",
            "Epoch 382 Loss 0.0408\n",
            "Epoch 383 Loss 0.0408\n",
            "Epoch 384 Loss 0.0408\n",
            "Epoch 385 Loss 0.0408\n",
            "Epoch 386 Loss 0.0408\n",
            "Epoch 387 Loss 0.0408\n",
            "Epoch 388 Loss 0.0407\n",
            "Epoch 389 Loss 0.0407\n",
            "Epoch 390 Loss 0.0407\n",
            "Epoch 391 Loss 0.0407\n",
            "Epoch 392 Loss 0.0407\n",
            "Epoch 393 Loss 0.0407\n",
            "Epoch 394 Loss 0.0407\n",
            "Epoch 395 Loss 0.0406\n",
            "Epoch 396 Loss 0.0406\n",
            "Epoch 397 Loss 0.0406\n",
            "Epoch 398 Loss 0.0406\n",
            "Epoch 399 Loss 0.0406\n",
            "Epoch 400 Loss 0.0406\n",
            "Epoch 401 Loss 0.0406\n",
            "Epoch 402 Loss 0.0405\n",
            "Epoch 403 Loss 0.0405\n",
            "Epoch 404 Loss 0.0405\n",
            "Epoch 405 Loss 0.0405\n",
            "Epoch 406 Loss 0.0405\n",
            "Epoch 407 Loss 0.0405\n",
            "Epoch 408 Loss 0.0405\n",
            "Epoch 409 Loss 0.0404\n",
            "Epoch 410 Loss 0.0404\n",
            "Epoch 411 Loss 0.0404\n",
            "Epoch 412 Loss 0.0404\n",
            "Epoch 413 Loss 0.0404\n",
            "Epoch 414 Loss 0.0404\n",
            "Epoch 415 Loss 0.0404\n",
            "Epoch 416 Loss 0.0403\n",
            "Epoch 417 Loss 0.0403\n",
            "Epoch 418 Loss 0.0403\n",
            "Epoch 419 Loss 0.0403\n",
            "Epoch 420 Loss 0.0403\n",
            "Epoch 421 Loss 0.0403\n",
            "Epoch 422 Loss 0.0403\n",
            "Epoch 423 Loss 0.0403\n",
            "Epoch 424 Loss 0.0402\n",
            "Epoch 425 Loss 0.0402\n",
            "Epoch 426 Loss 0.0402\n",
            "Epoch 427 Loss 0.0402\n",
            "Epoch 428 Loss 0.0402\n",
            "Epoch 429 Loss 0.0402\n",
            "Epoch 430 Loss 0.0402\n",
            "Epoch 431 Loss 0.0402\n",
            "Epoch 432 Loss 0.0401\n",
            "Epoch 433 Loss 0.0401\n",
            "Epoch 434 Loss 0.0401\n",
            "Epoch 435 Loss 0.0401\n",
            "Epoch 436 Loss 0.0401\n",
            "Epoch 437 Loss 0.0401\n",
            "Epoch 438 Loss 0.0401\n",
            "Epoch 439 Loss 0.0401\n",
            "Epoch 440 Loss 0.0400\n",
            "Epoch 441 Loss 0.0400\n",
            "Epoch 442 Loss 0.0400\n",
            "Epoch 443 Loss 0.0400\n",
            "Epoch 444 Loss 0.0400\n",
            "Epoch 445 Loss 0.0400\n",
            "Epoch 446 Loss 0.0400\n",
            "Epoch 447 Loss 0.0400\n",
            "Epoch 448 Loss 0.0399\n",
            "Epoch 449 Loss 0.0399\n",
            "Epoch 450 Loss 0.0399\n",
            "Epoch 451 Loss 0.0399\n",
            "Epoch 452 Loss 0.0399\n",
            "Epoch 453 Loss 0.0399\n",
            "Epoch 454 Loss 0.0399\n",
            "Epoch 455 Loss 0.0399\n",
            "Epoch 456 Loss 0.0399\n",
            "Epoch 457 Loss 0.0398\n",
            "Epoch 458 Loss 0.0398\n",
            "Epoch 459 Loss 0.0398\n",
            "Epoch 460 Loss 0.0398\n",
            "Epoch 461 Loss 0.0398\n",
            "Epoch 462 Loss 0.0398\n",
            "Epoch 463 Loss 0.0398\n",
            "Epoch 464 Loss 0.0398\n",
            "Epoch 465 Loss 0.0397\n",
            "Epoch 466 Loss 0.0397\n",
            "Epoch 467 Loss 0.0397\n",
            "Epoch 468 Loss 0.0397\n",
            "Epoch 469 Loss 0.0397\n",
            "Epoch 470 Loss 0.0397\n",
            "Epoch 471 Loss 0.0397\n",
            "Epoch 472 Loss 0.0397\n",
            "Epoch 473 Loss 0.0397\n",
            "Epoch 474 Loss 0.0396\n",
            "Epoch 475 Loss 0.0396\n",
            "Epoch 476 Loss 0.0396\n",
            "Epoch 477 Loss 0.0396\n",
            "Epoch 478 Loss 0.0396\n",
            "Epoch 479 Loss 0.0396\n",
            "Epoch 480 Loss 0.0396\n",
            "Epoch 481 Loss 0.0396\n",
            "Epoch 482 Loss 0.0396\n",
            "Epoch 483 Loss 0.0396\n",
            "Epoch 484 Loss 0.0395\n",
            "Epoch 485 Loss 0.0395\n",
            "Epoch 486 Loss 0.0395\n",
            "Epoch 487 Loss 0.0395\n",
            "Epoch 488 Loss 0.0395\n",
            "Epoch 489 Loss 0.0395\n",
            "Epoch 490 Loss 0.0395\n",
            "Epoch 491 Loss 0.0395\n",
            "Epoch 492 Loss 0.0395\n",
            "Epoch 493 Loss 0.0394\n",
            "Epoch 494 Loss 0.0394\n",
            "Epoch 495 Loss 0.0394\n",
            "Epoch 496 Loss 0.0394\n",
            "Epoch 497 Loss 0.0394\n",
            "Epoch 498 Loss 0.0394\n",
            "Epoch 499 Loss 0.0394\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I5Mfx55VpH-V",
        "colab_type": "code",
        "outputId": "b71d6198-44a8-4837-918c-576624797668",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        }
      },
      "source": [
        "plt.plot(losses)\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.show()"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEGCAYAAAB/+QKOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmVlVDggzM6vKAWFmZlX9fz3NRJKEMibkAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}