titanic.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Gongsoo/Python3/blob/main/titanic.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "id": "PyapyyUpBQdc",
        "outputId": "925762f8-7d3b-4ae4-aff4-061eb8f23ec2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cd \"/content/drive/MyDrive/\""
      ],
      "metadata": {
        "id": "h6mb5B93Be-i",
        "outputId": "08da1cf6-017b-4ad0-e770-3aacabd45a2b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "/content/drive/MyDrive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!unzip -qq \"/content/drive/MyDrive/타이타닉.zip\""
      ],
      "metadata": {
        "id": "J5lqGt68CQii"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!mkdir titanic_pytorch\n"
      ],
      "metadata": {
        "id": "Z3pErZd7Cpbb"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import shutil\n",
        "shutil.move('/content/drive/MyDrive/train.csv','/content/drive/MyDrive/titanic_pytorch/train.csv')\n",
        "shutil.move('/content/drive/MyDrive/test.csv','/content/drive/MyDrive/titanic_pytorch/test.csv')\n",
        "shutil.move('/content/drive/MyDrive/submission.csv','/content/drive/MyDrive/titanic_pytorch/submission.csv')"
      ],
      "metadata": {
        "id": "UXuAtq95CtfZ",
        "outputId": "9c142583-9b13-4041-d359-53b4a48d740f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'/content/drive/MyDrive/titanic_pytorch/submission.csv'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "shutil.move('/content/drive/MyDrive/타이타닉.zip','/content/drive/MyDrive/titanic_pytorch/titanic_data.zip')"
      ],
      "metadata": {
        "id": "kEzOmeA2DV0s",
        "outputId": "7332b887-2c90-4729-f794-792ca59255bb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'/content/drive/MyDrive/titanic_pytorch/titanic_data.zip'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import numpy as np"
      ],
      "metadata": {
        "id": "1WWupEEgDhyC"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train = pd.read_csv('/content/drive/MyDrive/titanic_pytorch/train.csv')\n",
        "test = pd.read_csv('/content/drive/MyDrive/titanic_pytorch/test.csv')\n",
        "sub = pd.read_csv('/content/drive/MyDrive/titanic_pytorch/submission.csv')"
      ],
      "metadata": {
        "id": "rbYArzUOISIR"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train = train.drop('Ticket',axis = 1)\n",
        "train = train.drop('Cabin',axis = 1)\n",
        "test = test.drop('Ticket',axis = 1)\n",
        "test = test.drop('Cabin',axis = 1)"
      ],
      "metadata": {
        "id": "xKfJrfA_Iw3F"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train = train.fillna({'Embarked':'S'})\n",
        "test = test.fillna({'Embarked':'S','Fare':20})"
      ],
      "metadata": {
        "id": "DtjoMOtfJMb9"
      },
      "execution_count": 71,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "em_map = {'S':0,'Q':1,'C':2}\n",
        "\n",
        "for t in [train] :\n",
        "  t['Embarked'] = t['Embarked'].map(em_map)\n",
        "for t in [test] :\n",
        "  t['Embarked'] = t['Embarked'].map(em_map)"
      ],
      "metadata": {
        "id": "4WWyRmQaKqGk"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "s_map = {'male':1, 'female':2}\n",
        "\n",
        "for t in [train] :\n",
        "  t['Sex'] = t['Sex'].map(s_map)\n",
        "for t in [test] :\n",
        "  t['Sex'] = t['Sex'].map(s_map)"
      ],
      "metadata": {
        "id": "kZUxK3idLPFd"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "age_map = {'Mr.':20,'Miss.':15,'Mrs.':20,'Ms.':10,'Mme.':30,'Mlle.':15,'Dona.':30,'Countess.':30,'Sir.':25,'Lady.':20,'Jonkheer.':20,'Don.':20,'Col.':35,'Major.':30,'Capt.':23,'Dr.':30,'Master.':25,'Rev.':35}\n",
        "\n",
        "for t in [train] :\n",
        "  for i in range(len(t['Age'])) :\n",
        "    if np.isnan(t['Age'][i]) :\n",
        "      for j in age_map :\n",
        "        if j in t['Name'][i] :\n",
        "          t['Age'][i] = age_map[j]\n",
        "for t in [test] :\n",
        "  for i in range(len(t['Age'])) :\n",
        "    if np.isnan(t['Age'][i]) :\n",
        "      for j in age_map :\n",
        "        if j in t['Name'][i] :\n",
        "          t['Age'][i] = age_map[j]"
      ],
      "metadata": {
        "id": "YRcFZNHSOqBw",
        "outputId": "f4ac0e87-9564-4ad0-ea81-66489fa0e1b6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-67-69fb9fd690b4>:14: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  t['Age'][i] = age_map[j]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train = train.drop(['Name'], axis=1)\n",
        "test = test.drop(['Name'], axis=1)"
      ],
      "metadata": {
        "id": "_SCFRED1-0E7"
      },
      "execution_count": 73,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train = train.drop(['PassengerId'], axis=1)\n",
        "test = test.drop(['PassengerId'], axis=1)"
      ],
      "metadata": {
        "id": "4xfCYzajDtdN"
      },
      "execution_count": 74,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x = train.drop(['Survived'],axis=1)\n",
        "y = train['Survived']"
      ],
      "metadata": {
        "id": "1cVbqMvUDtiq",
        "outputId": "f0786d24-f160-4550-9092-ae384c884f17",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 891 entries, 0 to 890\n",
            "Data columns (total 7 columns):\n",
            " #   Column    Non-Null Count  Dtype  \n",
            "---  ------    --------------  -----  \n",
            " 0   Pclass    891 non-null    int64  \n",
            " 1   Sex       891 non-null    int64  \n",
            " 2   Age       891 non-null    float64\n",
            " 3   SibSp     891 non-null    int64  \n",
            " 4   Parch     891 non-null    int64  \n",
            " 5   Fare      891 non-null    float64\n",
            " 6   Embarked  891 non-null    int64  \n",
            "dtypes: float64(2), int64(5)\n",
            "memory usage: 48.9 KB\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    0\n",
              "1    1\n",
              "2    1\n",
              "3    1\n",
              "4    0\n",
              "Name: Survived, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "import torch\n",
        "from torch import nn\n",
        "from torch.utils.data import DataLoader"
      ],
      "metadata": {
        "id": "1Dj_WV3zJXR3"
      },
      "execution_count": 80,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 정규화? 이거왜함?\n",
        "\n",
        "scaler = StandardScaler()\n",
        "scaler.fit(x)\n",
        "x_scaled = scaler.transform(x)\n",
        "\n",
        "scaler.fit(test)\n",
        "test_scaled = scaler.transform(test)"
      ],
      "metadata": {
        "id": "f-FNut6ZEzr8"
      },
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(y)"
      ],
      "metadata": {
        "id": "esa-aEbUKmYM",
        "outputId": "97328012-77eb-4d50-bf0f-95235ea9d7b2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0      0\n",
            "1      1\n",
            "2      1\n",
            "3      1\n",
            "4      0\n",
            "      ..\n",
            "886    0\n",
            "887    1\n",
            "888    0\n",
            "889    1\n",
            "890    0\n",
            "Name: Survived, Length: 891, dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_x = torch.tensor(x_scaled).float()\n",
        "train_y = torch.tensor(y).resize(891).float()\n",
        "test_x = torch.tensor(test.values).float()\n",
        "\n",
        "model = torch.nn.Sequential(\n",
        "    torch.nn.Linear(7,1),\n",
        "    torch.nn.Sigmoid()\n",
        ")\n",
        "\n",
        "optimizer = torch.optim.Adam(model.parameters(),lr=0.00001)\n",
        "epochs = 15000"
      ],
      "metadata": {
        "id": "JuKMbW5lJ_Sc",
        "outputId": "563591bb-f9b4-4791-ba00-ee78101f622c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 128,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/torch/_tensor.py:761: UserWarning: non-inplace resize is deprecated\n",
            "  warnings.warn(\"non-inplace resize is deprecated\")\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for epoch in range(epochs) :\n",
        "  pred = model(train_x)\n",
        "  pred = pred.view(-1)\n",
        "  error = torch.nn.functional.binary_cross_entropy(pred,train_y)\n",
        "\n",
        "  optimizer.zero_grad()\n",
        "  error.backward()\n",
        "  optimizer.step()\n",
        "\n",
        "  if epoch%10==0 :\n",
        "    \n",
        "    output = pred >= torch.FloatTensor([0.5])\n",
        "    correct = output.float() == train_y\n",
        "    acc = correct.sum().item()/len(correct)\n",
        "    print('Epoch {:4d}/{} Cost: {:.6f} Accuracy {:2.2f}%'.format(epoch,epochs,error.item(),acc*100))"
      ],
      "metadata": {
        "id": "EjdSZo2TNhL9",
        "outputId": "8d539083-0983-4ac3-a181-2787ba01300c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 129,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch    0/15000 Cost: 0.744567 Accuracy 42.42%\n",
            "Epoch   10/15000 Cost: 0.744478 Accuracy 42.42%\n",
            "Epoch   20/15000 Cost: 0.744388 Accuracy 42.42%\n",
            "Epoch   30/15000 Cost: 0.744300 Accuracy 42.42%\n",
            "Epoch   40/15000 Cost: 0.744210 Accuracy 42.42%\n",
            "Epoch   50/15000 Cost: 0.744121 Accuracy 42.42%\n",
            "Epoch   60/15000 Cost: 0.744033 Accuracy 42.42%\n",
            "Epoch   70/15000 Cost: 0.743944 Accuracy 42.31%\n",
            "Epoch   80/15000 Cost: 0.743855 Accuracy 42.31%\n",
            "Epoch   90/15000 Cost: 0.743766 Accuracy 42.31%\n",
            "Epoch  100/15000 Cost: 0.743677 Accuracy 42.31%\n",
            "Epoch  110/15000 Cost: 0.743588 Accuracy 42.31%\n",
            "Epoch  120/15000 Cost: 0.743500 Accuracy 42.31%\n",
            "Epoch  130/15000 Cost: 0.743411 Accuracy 42.31%\n",
            "Epoch  140/15000 Cost: 0.743322 Accuracy 42.31%\n",
            "Epoch  150/15000 Cost: 0.743234 Accuracy 42.31%\n",
            "Epoch  160/15000 Cost: 0.743145 Accuracy 42.31%\n",
            "Epoch  170/15000 Cost: 0.743057 Accuracy 42.31%\n",
            "Epoch  180/15000 Cost: 0.742968 Accuracy 42.31%\n",
            "Epoch  190/15000 Cost: 0.742880 Accuracy 42.31%\n",
            "Epoch  200/15000 Cost: 0.742791 Accuracy 42.31%\n",
            "Epoch  210/15000 Cost: 0.742703 Accuracy 42.31%\n",
            "Epoch  220/15000 Cost: 0.742615 Accuracy 42.31%\n",
            "Epoch  230/15000 Cost: 0.742527 Accuracy 42.31%\n",
            "Epoch  240/15000 Cost: 0.742438 Accuracy 42.31%\n",
            "Epoch  250/15000 Cost: 0.742350 Accuracy 42.31%\n",
            "Epoch  260/15000 Cost: 0.742262 Accuracy 42.31%\n",
            "Epoch  270/15000 Cost: 0.742174 Accuracy 42.31%\n",
            "Epoch  280/15000 Cost: 0.742086 Accuracy 42.31%\n",
            "Epoch  290/15000 Cost: 0.741997 Accuracy 42.42%\n",
            "Epoch  300/15000 Cost: 0.741910 Accuracy 42.42%\n",
            "Epoch  310/15000 Cost: 0.741821 Accuracy 42.54%\n",
            "Epoch  320/15000 Cost: 0.741733 Accuracy 42.54%\n",
            "Epoch  330/15000 Cost: 0.741646 Accuracy 42.54%\n",
            "Epoch  340/15000 Cost: 0.741558 Accuracy 42.42%\n",
            "Epoch  350/15000 Cost: 0.741470 Accuracy 42.42%\n",
            "Epoch  360/15000 Cost: 0.741382 Accuracy 42.42%\n",
            "Epoch  370/15000 Cost: 0.741294 Accuracy 42.42%\n",
            "Epoch  380/15000 Cost: 0.741206 Accuracy 42.42%\n",
            "Epoch  390/15000 Cost: 0.741119 Accuracy 42.42%\n",
            "Epoch  400/15000 Cost: 0.741031 Accuracy 42.42%\n",
            "Epoch  410/15000 Cost: 0.740943 Accuracy 42.42%\n",
            "Epoch  420/15000 Cost: 0.740856 Accuracy 42.42%\n",
            "Epoch  430/15000 Cost: 0.740768 Accuracy 42.42%\n",
            "Epoch  440/15000 Cost: 0.740680 Accuracy 42.42%\n",
            "Epoch  450/15000 Cost: 0.740593 Accuracy 42.42%\n",
            "Epoch  460/15000 Cost: 0.740505 Accuracy 42.42%\n",
            "Epoch  470/15000 Cost: 0.740418 Accuracy 42.42%\n",
            "Epoch  480/15000 Cost: 0.740331 Accuracy 42.54%\n",
            "Epoch  490/15000 Cost: 0.740243 Accuracy 42.54%\n",
            "Epoch  500/15000 Cost: 0.740156 Accuracy 42.54%\n",
            "Epoch  510/15000 Cost: 0.740069 Accuracy 42.54%\n",
            "Epoch  520/15000 Cost: 0.739981 Accuracy 42.54%\n",
            "Epoch  530/15000 Cost: 0.739894 Accuracy 42.54%\n",
            "Epoch  540/15000 Cost: 0.739807 Accuracy 42.54%\n",
            "Epoch  550/15000 Cost: 0.739720 Accuracy 42.54%\n",
            "Epoch  560/15000 Cost: 0.739633 Accuracy 42.54%\n",
            "Epoch  570/15000 Cost: 0.739546 Accuracy 42.54%\n",
            "Epoch  580/15000 Cost: 0.739459 Accuracy 42.54%\n",
            "Epoch  590/15000 Cost: 0.739372 Accuracy 42.54%\n",
            "Epoch  600/15000 Cost: 0.739285 Accuracy 42.54%\n",
            "Epoch  610/15000 Cost: 0.739198 Accuracy 42.54%\n",
            "Epoch  620/15000 Cost: 0.739111 Accuracy 42.54%\n",
            "Epoch  630/15000 Cost: 0.739024 Accuracy 42.54%\n",
            "Epoch  640/15000 Cost: 0.738937 Accuracy 42.54%\n",
            "Epoch  650/15000 Cost: 0.738850 Accuracy 42.54%\n",
            "Epoch  660/15000 Cost: 0.738763 Accuracy 42.54%\n",
            "Epoch  670/15000 Cost: 0.738677 Accuracy 42.54%\n",
            "Epoch  680/15000 Cost: 0.738590 Accuracy 42.54%\n",
            "Epoch  690/15000 Cost: 0.738503 Accuracy 42.54%\n",
            "Epoch  700/15000 Cost: 0.738416 Accuracy 42.42%\n",
            "Epoch  710/15000 Cost: 0.738330 Accuracy 42.42%\n",
            "Epoch  720/15000 Cost: 0.738243 Accuracy 42.42%\n",
            "Epoch  730/15000 Cost: 0.738157 Accuracy 42.42%\n",
            "Epoch  740/15000 Cost: 0.738070 Accuracy 42.42%\n",
            "Epoch  750/15000 Cost: 0.737984 Accuracy 42.42%\n",
            "Epoch  760/15000 Cost: 0.737897 Accuracy 42.54%\n",
            "Epoch  770/15000 Cost: 0.737811 Accuracy 42.54%\n",
            "Epoch  780/15000 Cost: 0.737725 Accuracy 42.54%\n",
            "Epoch  790/15000 Cost: 0.737638 Accuracy 42.54%\n",
            "Epoch  800/15000 Cost: 0.737552 Accuracy 42.54%\n",
            "Epoch  810/15000 Cost: 0.737466 Accuracy 42.54%\n",
            "Epoch  820/15000 Cost: 0.737379 Accuracy 42.54%\n",
            "Epoch  830/15000 Cost: 0.737293 Accuracy 42.54%\n",
            "Epoch  840/15000 Cost: 0.737207 Accuracy 42.54%\n",
            "Epoch  850/15000 Cost: 0.737121 Accuracy 42.54%\n",
            "Epoch  860/15000 Cost: 0.737035 Accuracy 42.54%\n",
            "Epoch  870/15000 Cost: 0.736949 Accuracy 42.65%\n",
            "Epoch  880/15000 Cost: 0.736863 Accuracy 42.65%\n",
            "Epoch  890/15000 Cost: 0.736777 Accuracy 42.65%\n",
            "Epoch  900/15000 Cost: 0.736691 Accuracy 42.65%\n",
            "Epoch  910/15000 Cost: 0.736605 Accuracy 42.65%\n",
            "Epoch  920/15000 Cost: 0.736519 Accuracy 42.76%\n",
            "Epoch  930/15000 Cost: 0.736433 Accuracy 42.76%\n",
            "Epoch  940/15000 Cost: 0.736347 Accuracy 42.76%\n",
            "Epoch  950/15000 Cost: 0.736261 Accuracy 42.87%\n",
            "Epoch  960/15000 Cost: 0.736176 Accuracy 42.87%\n",
            "Epoch  970/15000 Cost: 0.736090 Accuracy 42.99%\n",
            "Epoch  980/15000 Cost: 0.736004 Accuracy 42.99%\n",
            "Epoch  990/15000 Cost: 0.735918 Accuracy 42.99%\n",
            "Epoch 1000/15000 Cost: 0.735833 Accuracy 43.10%\n",
            "Epoch 1010/15000 Cost: 0.735747 Accuracy 43.10%\n",
            "Epoch 1020/15000 Cost: 0.735662 Accuracy 43.10%\n",
            "Epoch 1030/15000 Cost: 0.735576 Accuracy 43.10%\n",
            "Epoch 1040/15000 Cost: 0.735491 Accuracy 43.10%\n",
            "Epoch 1050/15000 Cost: 0.735405 Accuracy 43.10%\n",
            "Epoch 1060/15000 Cost: 0.735320 Accuracy 43.10%\n",
            "Epoch 1070/15000 Cost: 0.735234 Accuracy 43.10%\n",
            "Epoch 1080/15000 Cost: 0.735149 Accuracy 43.10%\n",
            "Epoch 1090/15000 Cost: 0.735063 Accuracy 43.10%\n",
            "Epoch 1100/15000 Cost: 0.734978 Accuracy 43.10%\n",
            "Epoch 1110/15000 Cost: 0.734893 Accuracy 43.21%\n",
            "Epoch 1120/15000 Cost: 0.734807 Accuracy 43.21%\n",
            "Epoch 1130/15000 Cost: 0.734722 Accuracy 43.21%\n",
            "Epoch 1140/15000 Cost: 0.734637 Accuracy 43.21%\n",
            "Epoch 1150/15000 Cost: 0.734552 Accuracy 43.43%\n",
            "Epoch 1160/15000 Cost: 0.734467 Accuracy 43.43%\n",
            "Epoch 1170/15000 Cost: 0.734382 Accuracy 43.43%\n",
            "Epoch 1180/15000 Cost: 0.734297 Accuracy 43.43%\n",
            "Epoch 1190/15000 Cost: 0.734212 Accuracy 43.43%\n",
            "Epoch 1200/15000 Cost: 0.734127 Accuracy 43.43%\n",
            "Epoch 1210/15000 Cost: 0.734042 Accuracy 43.43%\n",
            "Epoch 1220/15000 Cost: 0.733957 Accuracy 43.43%\n",
            "Epoch 1230/15000 Cost: 0.733872 Accuracy 43.43%\n",
            "Epoch 1240/15000 Cost: 0.733787 Accuracy 43.43%\n",
            "Epoch 1250/15000 Cost: 0.733702 Accuracy 43.43%\n",
            "Epoch 1260/15000 Cost: 0.733617 Accuracy 43.43%\n",
            "Epoch 1270/15000 Cost: 0.733533 Accuracy 43.43%\n",
            "Epoch 1280/15000 Cost: 0.733448 Accuracy 43.55%\n",
            "Epoch 1290/15000 Cost: 0.733363 Accuracy 43.55%\n",
            "Epoch 1300/15000 Cost: 0.733279 Accuracy 43.55%\n",
            "Epoch 1310/15000 Cost: 0.733194 Accuracy 43.55%\n",
            "Epoch 1320/15000 Cost: 0.733109 Accuracy 43.55%\n",
            "Epoch 1330/15000 Cost: 0.733025 Accuracy 43.55%\n",
            "Epoch 1340/15000 Cost: 0.732940 Accuracy 43.55%\n",
            "Epoch 1350/15000 Cost: 0.732856 Accuracy 43.55%\n",
            "Epoch 1360/15000 Cost: 0.732771 Accuracy 43.55%\n",
            "Epoch 1370/15000 Cost: 0.732687 Accuracy 43.55%\n",
            "Epoch 1380/15000 Cost: 0.732602 Accuracy 43.55%\n",
            "Epoch 1390/15000 Cost: 0.732518 Accuracy 43.55%\n",
            "Epoch 1400/15000 Cost: 0.732433 Accuracy 43.55%\n",
            "Epoch 1410/15000 Cost: 0.732349 Accuracy 43.55%\n",
            "Epoch 1420/15000 Cost: 0.732265 Accuracy 43.55%\n",
            "Epoch 1430/15000 Cost: 0.732181 Accuracy 43.55%\n",
            "Epoch 1440/15000 Cost: 0.732096 Accuracy 43.66%\n",
            "Epoch 1450/15000 Cost: 0.732012 Accuracy 43.66%\n",
            "Epoch 1460/15000 Cost: 0.731928 Accuracy 43.66%\n",
            "Epoch 1470/15000 Cost: 0.731844 Accuracy 43.55%\n",
            "Epoch 1480/15000 Cost: 0.731760 Accuracy 43.55%\n",
            "Epoch 1490/15000 Cost: 0.731676 Accuracy 43.55%\n",
            "Epoch 1500/15000 Cost: 0.731591 Accuracy 43.55%\n",
            "Epoch 1510/15000 Cost: 0.731507 Accuracy 43.66%\n",
            "Epoch 1520/15000 Cost: 0.731423 Accuracy 43.66%\n",
            "Epoch 1530/15000 Cost: 0.731339 Accuracy 43.66%\n",
            "Epoch 1540/15000 Cost: 0.731255 Accuracy 43.66%\n",
            "Epoch 1550/15000 Cost: 0.731172 Accuracy 43.66%\n",
            "Epoch 1560/15000 Cost: 0.731088 Accuracy 43.66%\n",
            "Epoch 1570/15000 Cost: 0.731004 Accuracy 43.66%\n",
            "Epoch 1580/15000 Cost: 0.730920 Accuracy 43.77%\n",
            "Epoch 1590/15000 Cost: 0.730836 Accuracy 43.77%\n",
            "Epoch 1600/15000 Cost: 0.730753 Accuracy 43.77%\n",
            "Epoch 1610/15000 Cost: 0.730669 Accuracy 43.77%\n",
            "Epoch 1620/15000 Cost: 0.730585 Accuracy 43.77%\n",
            "Epoch 1630/15000 Cost: 0.730501 Accuracy 43.77%\n",
            "Epoch 1640/15000 Cost: 0.730418 Accuracy 43.77%\n",
            "Epoch 1650/15000 Cost: 0.730334 Accuracy 43.77%\n",
            "Epoch 1660/15000 Cost: 0.730251 Accuracy 43.77%\n",
            "Epoch 1670/15000 Cost: 0.730167 Accuracy 43.77%\n",
            "Epoch 1680/15000 Cost: 0.730084 Accuracy 43.77%\n",
            "Epoch 1690/15000 Cost: 0.730000 Accuracy 43.77%\n",
            "Epoch 1700/15000 Cost: 0.729917 Accuracy 43.77%\n",
            "Epoch 1710/15000 Cost: 0.729833 Accuracy 43.66%\n",
            "Epoch 1720/15000 Cost: 0.729750 Accuracy 43.66%\n",
            "Epoch 1730/15000 Cost: 0.729667 Accuracy 43.66%\n",
            "Epoch 1740/15000 Cost: 0.729583 Accuracy 43.66%\n",
            "Epoch 1750/15000 Cost: 0.729500 Accuracy 43.66%\n",
            "Epoch 1760/15000 Cost: 0.729417 Accuracy 43.66%\n",
            "Epoch 1770/15000 Cost: 0.729334 Accuracy 43.66%\n",
            "Epoch 1780/15000 Cost: 0.729250 Accuracy 43.66%\n",
            "Epoch 1790/15000 Cost: 0.729167 Accuracy 43.66%\n",
            "Epoch 1800/15000 Cost: 0.729084 Accuracy 43.66%\n",
            "Epoch 1810/15000 Cost: 0.729001 Accuracy 43.66%\n",
            "Epoch 1820/15000 Cost: 0.728918 Accuracy 43.66%\n",
            "Epoch 1830/15000 Cost: 0.728835 Accuracy 43.66%\n",
            "Epoch 1840/15000 Cost: 0.728752 Accuracy 43.66%\n",
            "Epoch 1850/15000 Cost: 0.728669 Accuracy 43.88%\n",
            "Epoch 1860/15000 Cost: 0.728586 Accuracy 43.88%\n",
            "Epoch 1870/15000 Cost: 0.728503 Accuracy 44.00%\n",
            "Epoch 1880/15000 Cost: 0.728420 Accuracy 44.11%\n",
            "Epoch 1890/15000 Cost: 0.728337 Accuracy 44.11%\n",
            "Epoch 1900/15000 Cost: 0.728254 Accuracy 44.11%\n",
            "Epoch 1910/15000 Cost: 0.728171 Accuracy 44.22%\n",
            "Epoch 1920/15000 Cost: 0.728088 Accuracy 44.22%\n",
            "Epoch 1930/15000 Cost: 0.728006 Accuracy 44.11%\n",
            "Epoch 1940/15000 Cost: 0.727923 Accuracy 44.11%\n",
            "Epoch 1950/15000 Cost: 0.727840 Accuracy 44.11%\n",
            "Epoch 1960/15000 Cost: 0.727758 Accuracy 44.11%\n",
            "Epoch 1970/15000 Cost: 0.727675 Accuracy 44.11%\n",
            "Epoch 1980/15000 Cost: 0.727592 Accuracy 44.11%\n",
            "Epoch 1990/15000 Cost: 0.727510 Accuracy 44.11%\n",
            "Epoch 2000/15000 Cost: 0.727427 Accuracy 44.11%\n",
            "Epoch 2010/15000 Cost: 0.727345 Accuracy 44.11%\n",
            "Epoch 2020/15000 Cost: 0.727262 Accuracy 44.11%\n",
            "Epoch 2030/15000 Cost: 0.727180 Accuracy 44.11%\n",
            "Epoch 2040/15000 Cost: 0.727097 Accuracy 44.11%\n",
            "Epoch 2050/15000 Cost: 0.727015 Accuracy 44.11%\n",
            "Epoch 2060/15000 Cost: 0.726932 Accuracy 44.11%\n",
            "Epoch 2070/15000 Cost: 0.726850 Accuracy 44.11%\n",
            "Epoch 2080/15000 Cost: 0.726768 Accuracy 44.11%\n",
            "Epoch 2090/15000 Cost: 0.726685 Accuracy 44.11%\n",
            "Epoch 2100/15000 Cost: 0.726603 Accuracy 44.11%\n",
            "Epoch 2110/15000 Cost: 0.726521 Accuracy 44.11%\n",
            "Epoch 2120/15000 Cost: 0.726439 Accuracy 44.11%\n",
            "Epoch 2130/15000 Cost: 0.726356 Accuracy 44.11%\n",
            "Epoch 2140/15000 Cost: 0.726274 Accuracy 44.22%\n",
            "Epoch 2150/15000 Cost: 0.726192 Accuracy 44.22%\n",
            "Epoch 2160/15000 Cost: 0.726110 Accuracy 44.22%\n",
            "Epoch 2170/15000 Cost: 0.726028 Accuracy 44.22%\n",
            "Epoch 2180/15000 Cost: 0.725946 Accuracy 44.22%\n",
            "Epoch 2190/15000 Cost: 0.725864 Accuracy 44.22%\n",
            "Epoch 2200/15000 Cost: 0.725782 Accuracy 44.22%\n",
            "Epoch 2210/15000 Cost: 0.725700 Accuracy 44.22%\n",
            "Epoch 2220/15000 Cost: 0.725618 Accuracy 44.33%\n",
            "Epoch 2230/15000 Cost: 0.725536 Accuracy 44.22%\n",
            "Epoch 2240/15000 Cost: 0.725455 Accuracy 44.22%\n",
            "Epoch 2250/15000 Cost: 0.725373 Accuracy 44.22%\n",
            "Epoch 2260/15000 Cost: 0.725291 Accuracy 44.22%\n",
            "Epoch 2270/15000 Cost: 0.725209 Accuracy 44.22%\n",
            "Epoch 2280/15000 Cost: 0.725127 Accuracy 44.22%\n",
            "Epoch 2290/15000 Cost: 0.725046 Accuracy 44.22%\n",
            "Epoch 2300/15000 Cost: 0.724964 Accuracy 44.22%\n",
            "Epoch 2310/15000 Cost: 0.724882 Accuracy 44.22%\n",
            "Epoch 2320/15000 Cost: 0.724801 Accuracy 44.22%\n",
            "Epoch 2330/15000 Cost: 0.724719 Accuracy 44.22%\n",
            "Epoch 2340/15000 Cost: 0.724637 Accuracy 44.22%\n",
            "Epoch 2350/15000 Cost: 0.724556 Accuracy 44.22%\n",
            "Epoch 2360/15000 Cost: 0.724474 Accuracy 44.22%\n",
            "Epoch 2370/15000 Cost: 0.724393 Accuracy 44.22%\n",
            "Epoch 2380/15000 Cost: 0.724311 Accuracy 44.22%\n",
            "Epoch 2390/15000 Cost: 0.724230 Accuracy 44.22%\n",
            "Epoch 2400/15000 Cost: 0.724149 Accuracy 44.44%\n",
            "Epoch 2410/15000 Cost: 0.724067 Accuracy 44.44%\n",
            "Epoch 2420/15000 Cost: 0.723986 Accuracy 44.44%\n",
            "Epoch 2430/15000 Cost: 0.723904 Accuracy 44.44%\n",
            "Epoch 2440/15000 Cost: 0.723823 Accuracy 44.44%\n",
            "Epoch 2450/15000 Cost: 0.723742 Accuracy 44.44%\n",
            "Epoch 2460/15000 Cost: 0.723661 Accuracy 44.44%\n",
            "Epoch 2470/15000 Cost: 0.723579 Accuracy 44.44%\n",
            "Epoch 2480/15000 Cost: 0.723498 Accuracy 44.44%\n",
            "Epoch 2490/15000 Cost: 0.723417 Accuracy 44.44%\n",
            "Epoch 2500/15000 Cost: 0.723336 Accuracy 44.44%\n",
            "Epoch 2510/15000 Cost: 0.723255 Accuracy 44.44%\n",
            "Epoch 2520/15000 Cost: 0.723174 Accuracy 44.44%\n",
            "Epoch 2530/15000 Cost: 0.723093 Accuracy 44.44%\n",
            "Epoch 2540/15000 Cost: 0.723012 Accuracy 44.44%\n",
            "Epoch 2550/15000 Cost: 0.722931 Accuracy 44.44%\n",
            "Epoch 2560/15000 Cost: 0.722850 Accuracy 44.44%\n",
            "Epoch 2570/15000 Cost: 0.722769 Accuracy 44.56%\n",
            "Epoch 2580/15000 Cost: 0.722688 Accuracy 44.67%\n",
            "Epoch 2590/15000 Cost: 0.722607 Accuracy 44.67%\n",
            "Epoch 2600/15000 Cost: 0.722526 Accuracy 45.01%\n",
            "Epoch 2610/15000 Cost: 0.722445 Accuracy 45.01%\n",
            "Epoch 2620/15000 Cost: 0.722364 Accuracy 45.01%\n",
            "Epoch 2630/15000 Cost: 0.722284 Accuracy 45.12%\n",
            "Epoch 2640/15000 Cost: 0.722203 Accuracy 45.12%\n",
            "Epoch 2650/15000 Cost: 0.722122 Accuracy 45.12%\n",
            "Epoch 2660/15000 Cost: 0.722041 Accuracy 45.23%\n",
            "Epoch 2670/15000 Cost: 0.721961 Accuracy 45.23%\n",
            "Epoch 2680/15000 Cost: 0.721880 Accuracy 45.23%\n",
            "Epoch 2690/15000 Cost: 0.721799 Accuracy 45.23%\n",
            "Epoch 2700/15000 Cost: 0.721719 Accuracy 45.23%\n",
            "Epoch 2710/15000 Cost: 0.721638 Accuracy 45.23%\n",
            "Epoch 2720/15000 Cost: 0.721558 Accuracy 45.23%\n",
            "Epoch 2730/15000 Cost: 0.721477 Accuracy 45.45%\n",
            "Epoch 2740/15000 Cost: 0.721397 Accuracy 45.45%\n",
            "Epoch 2750/15000 Cost: 0.721316 Accuracy 45.57%\n",
            "Epoch 2760/15000 Cost: 0.721236 Accuracy 45.57%\n",
            "Epoch 2770/15000 Cost: 0.721155 Accuracy 45.68%\n",
            "Epoch 2780/15000 Cost: 0.721075 Accuracy 45.68%\n",
            "Epoch 2790/15000 Cost: 0.720995 Accuracy 45.68%\n",
            "Epoch 2800/15000 Cost: 0.720914 Accuracy 45.68%\n",
            "Epoch 2810/15000 Cost: 0.720834 Accuracy 45.68%\n",
            "Epoch 2820/15000 Cost: 0.720754 Accuracy 45.79%\n",
            "Epoch 2830/15000 Cost: 0.720673 Accuracy 45.79%\n",
            "Epoch 2840/15000 Cost: 0.720593 Accuracy 45.79%\n",
            "Epoch 2850/15000 Cost: 0.720513 Accuracy 45.79%\n",
            "Epoch 2860/15000 Cost: 0.720433 Accuracy 45.79%\n",
            "Epoch 2870/15000 Cost: 0.720353 Accuracy 45.79%\n",
            "Epoch 2880/15000 Cost: 0.720273 Accuracy 45.79%\n",
            "Epoch 2890/15000 Cost: 0.720193 Accuracy 45.79%\n",
            "Epoch 2900/15000 Cost: 0.720113 Accuracy 45.68%\n",
            "Epoch 2910/15000 Cost: 0.720033 Accuracy 45.68%\n",
            "Epoch 2920/15000 Cost: 0.719953 Accuracy 45.68%\n",
            "Epoch 2930/15000 Cost: 0.719873 Accuracy 45.68%\n",
            "Epoch 2940/15000 Cost: 0.719793 Accuracy 45.79%\n",
            "Epoch 2950/15000 Cost: 0.719713 Accuracy 45.79%\n",
            "Epoch 2960/15000 Cost: 0.719633 Accuracy 46.02%\n",
            "Epoch 2970/15000 Cost: 0.719553 Accuracy 46.02%\n",
            "Epoch 2980/15000 Cost: 0.719474 Accuracy 46.02%\n",
            "Epoch 2990/15000 Cost: 0.719394 Accuracy 45.90%\n",
            "Epoch 3000/15000 Cost: 0.719314 Accuracy 45.90%\n",
            "Epoch 3010/15000 Cost: 0.719234 Accuracy 45.90%\n",
            "Epoch 3020/15000 Cost: 0.719154 Accuracy 45.90%\n",
            "Epoch 3030/15000 Cost: 0.719075 Accuracy 45.90%\n",
            "Epoch 3040/15000 Cost: 0.718995 Accuracy 45.90%\n",
            "Epoch 3050/15000 Cost: 0.718916 Accuracy 45.90%\n",
            "Epoch 3060/15000 Cost: 0.718836 Accuracy 45.90%\n",
            "Epoch 3070/15000 Cost: 0.718756 Accuracy 45.90%\n",
            "Epoch 3080/15000 Cost: 0.718677 Accuracy 45.90%\n",
            "Epoch 3090/15000 Cost: 0.718597 Accuracy 45.90%\n",
            "Epoch 3100/15000 Cost: 0.718518 Accuracy 46.02%\n",
            "Epoch 3110/15000 Cost: 0.718438 Accuracy 45.90%\n",
            "Epoch 3120/15000 Cost: 0.718359 Accuracy 45.90%\n",
            "Epoch 3130/15000 Cost: 0.718280 Accuracy 45.90%\n",
            "Epoch 3140/15000 Cost: 0.718200 Accuracy 45.90%\n",
            "Epoch 3150/15000 Cost: 0.718121 Accuracy 45.90%\n",
            "Epoch 3160/15000 Cost: 0.718041 Accuracy 45.90%\n",
            "Epoch 3170/15000 Cost: 0.717962 Accuracy 45.79%\n",
            "Epoch 3180/15000 Cost: 0.717883 Accuracy 45.79%\n",
            "Epoch 3190/15000 Cost: 0.717804 Accuracy 45.79%\n",
            "Epoch 3200/15000 Cost: 0.717724 Accuracy 45.79%\n",
            "Epoch 3210/15000 Cost: 0.717645 Accuracy 45.79%\n",
            "Epoch 3220/15000 Cost: 0.717566 Accuracy 45.79%\n",
            "Epoch 3230/15000 Cost: 0.717487 Accuracy 45.79%\n",
            "Epoch 3240/15000 Cost: 0.717408 Accuracy 45.79%\n",
            "Epoch 3250/15000 Cost: 0.717328 Accuracy 45.79%\n",
            "Epoch 3260/15000 Cost: 0.717249 Accuracy 45.79%\n",
            "Epoch 3270/15000 Cost: 0.717170 Accuracy 45.79%\n",
            "Epoch 3280/15000 Cost: 0.717091 Accuracy 45.79%\n",
            "Epoch 3290/15000 Cost: 0.717012 Accuracy 45.68%\n",
            "Epoch 3300/15000 Cost: 0.716933 Accuracy 45.57%\n",
            "Epoch 3310/15000 Cost: 0.716854 Accuracy 45.68%\n",
            "Epoch 3320/15000 Cost: 0.716775 Accuracy 45.68%\n",
            "Epoch 3330/15000 Cost: 0.716696 Accuracy 45.68%\n",
            "Epoch 3340/15000 Cost: 0.716618 Accuracy 45.79%\n",
            "Epoch 3350/15000 Cost: 0.716539 Accuracy 45.79%\n",
            "Epoch 3360/15000 Cost: 0.716460 Accuracy 45.90%\n",
            "Epoch 3370/15000 Cost: 0.716381 Accuracy 45.90%\n",
            "Epoch 3380/15000 Cost: 0.716302 Accuracy 45.90%\n",
            "Epoch 3390/15000 Cost: 0.716224 Accuracy 45.90%\n",
            "Epoch 3400/15000 Cost: 0.716145 Accuracy 45.90%\n",
            "Epoch 3410/15000 Cost: 0.716066 Accuracy 45.90%\n",
            "Epoch 3420/15000 Cost: 0.715988 Accuracy 45.90%\n",
            "Epoch 3430/15000 Cost: 0.715909 Accuracy 45.90%\n",
            "Epoch 3440/15000 Cost: 0.715830 Accuracy 45.90%\n",
            "Epoch 3450/15000 Cost: 0.715752 Accuracy 45.90%\n",
            "Epoch 3460/15000 Cost: 0.715673 Accuracy 45.90%\n",
            "Epoch 3470/15000 Cost: 0.715595 Accuracy 45.90%\n",
            "Epoch 3480/15000 Cost: 0.715516 Accuracy 45.79%\n",
            "Epoch 3490/15000 Cost: 0.715438 Accuracy 45.79%\n",
            "Epoch 3500/15000 Cost: 0.715359 Accuracy 45.79%\n",
            "Epoch 3510/15000 Cost: 0.715281 Accuracy 45.79%\n",
            "Epoch 3520/15000 Cost: 0.715203 Accuracy 45.79%\n",
            "Epoch 3530/15000 Cost: 0.715124 Accuracy 45.79%\n",
            "Epoch 3540/15000 Cost: 0.715046 Accuracy 45.79%\n",
            "Epoch 3550/15000 Cost: 0.714967 Accuracy 45.68%\n",
            "Epoch 3560/15000 Cost: 0.714889 Accuracy 45.68%\n",
            "Epoch 3570/15000 Cost: 0.714811 Accuracy 45.68%\n",
            "Epoch 3580/15000 Cost: 0.714733 Accuracy 45.68%\n",
            "Epoch 3590/15000 Cost: 0.714654 Accuracy 45.68%\n",
            "Epoch 3600/15000 Cost: 0.714576 Accuracy 45.57%\n",
            "Epoch 3610/15000 Cost: 0.714498 Accuracy 45.57%\n",
            "Epoch 3620/15000 Cost: 0.714420 Accuracy 45.57%\n",
            "Epoch 3630/15000 Cost: 0.714342 Accuracy 45.57%\n",
            "Epoch 3640/15000 Cost: 0.714264 Accuracy 45.57%\n",
            "Epoch 3650/15000 Cost: 0.714185 Accuracy 45.57%\n",
            "Epoch 3660/15000 Cost: 0.714107 Accuracy 45.68%\n",
            "Epoch 3670/15000 Cost: 0.714029 Accuracy 45.68%\n",
            "Epoch 3680/15000 Cost: 0.713951 Accuracy 45.68%\n",
            "Epoch 3690/15000 Cost: 0.713873 Accuracy 45.68%\n",
            "Epoch 3700/15000 Cost: 0.713795 Accuracy 45.68%\n",
            "Epoch 3710/15000 Cost: 0.713718 Accuracy 45.68%\n",
            "Epoch 3720/15000 Cost: 0.713640 Accuracy 45.68%\n",
            "Epoch 3730/15000 Cost: 0.713562 Accuracy 45.68%\n",
            "Epoch 3740/15000 Cost: 0.713484 Accuracy 45.68%\n",
            "Epoch 3750/15000 Cost: 0.713406 Accuracy 45.68%\n",
            "Epoch 3760/15000 Cost: 0.713328 Accuracy 45.68%\n",
            "Epoch 3770/15000 Cost: 0.713250 Accuracy 45.68%\n",
            "Epoch 3780/15000 Cost: 0.713173 Accuracy 45.68%\n",
            "Epoch 3790/15000 Cost: 0.713095 Accuracy 45.68%\n",
            "Epoch 3800/15000 Cost: 0.713017 Accuracy 45.68%\n",
            "Epoch 3810/15000 Cost: 0.712940 Accuracy 45.68%\n",
            "Epoch 3820/15000 Cost: 0.712862 Accuracy 45.68%\n",
            "Epoch 3830/15000 Cost: 0.712784 Accuracy 45.79%\n",
            "Epoch 3840/15000 Cost: 0.712707 Accuracy 45.79%\n",
            "Epoch 3850/15000 Cost: 0.712629 Accuracy 45.90%\n",
            "Epoch 3860/15000 Cost: 0.712552 Accuracy 45.90%\n",
            "Epoch 3870/15000 Cost: 0.712474 Accuracy 45.90%\n",
            "Epoch 3880/15000 Cost: 0.712397 Accuracy 45.90%\n",
            "Epoch 3890/15000 Cost: 0.712319 Accuracy 45.90%\n",
            "Epoch 3900/15000 Cost: 0.712242 Accuracy 45.90%\n",
            "Epoch 3910/15000 Cost: 0.712164 Accuracy 45.90%\n",
            "Epoch 3920/15000 Cost: 0.712087 Accuracy 46.02%\n",
            "Epoch 3930/15000 Cost: 0.712010 Accuracy 46.02%\n",
            "Epoch 3940/15000 Cost: 0.711932 Accuracy 46.02%\n",
            "Epoch 3950/15000 Cost: 0.711855 Accuracy 46.02%\n",
            "Epoch 3960/15000 Cost: 0.711778 Accuracy 46.02%\n",
            "Epoch 3970/15000 Cost: 0.711700 Accuracy 46.13%\n",
            "Epoch 3980/15000 Cost: 0.711623 Accuracy 46.13%\n",
            "Epoch 3990/15000 Cost: 0.711546 Accuracy 46.24%\n",
            "Epoch 4000/15000 Cost: 0.711469 Accuracy 46.46%\n",
            "Epoch 4010/15000 Cost: 0.711392 Accuracy 46.46%\n",
            "Epoch 4020/15000 Cost: 0.711314 Accuracy 46.46%\n",
            "Epoch 4030/15000 Cost: 0.711237 Accuracy 46.58%\n",
            "Epoch 4040/15000 Cost: 0.711160 Accuracy 46.58%\n",
            "Epoch 4050/15000 Cost: 0.711083 Accuracy 46.69%\n",
            "Epoch 4060/15000 Cost: 0.711006 Accuracy 46.69%\n",
            "Epoch 4070/15000 Cost: 0.710929 Accuracy 46.69%\n",
            "Epoch 4080/15000 Cost: 0.710852 Accuracy 46.69%\n",
            "Epoch 4090/15000 Cost: 0.710775 Accuracy 46.69%\n",
            "Epoch 4100/15000 Cost: 0.710698 Accuracy 46.80%\n",
            "Epoch 4110/15000 Cost: 0.710621 Accuracy 46.80%\n",
            "Epoch 4120/15000 Cost: 0.710544 Accuracy 46.80%\n",
            "Epoch 4130/15000 Cost: 0.710468 Accuracy 46.80%\n",
            "Epoch 4140/15000 Cost: 0.710391 Accuracy 46.80%\n",
            "Epoch 4150/15000 Cost: 0.710314 Accuracy 47.03%\n",
            "Epoch 4160/15000 Cost: 0.710237 Accuracy 46.91%\n",
            "Epoch 4170/15000 Cost: 0.710160 Accuracy 46.91%\n",
            "Epoch 4180/15000 Cost: 0.710084 Accuracy 46.91%\n",
            "Epoch 4190/15000 Cost: 0.710007 Accuracy 46.91%\n",
            "Epoch 4200/15000 Cost: 0.709930 Accuracy 46.91%\n",
            "Epoch 4210/15000 Cost: 0.709854 Accuracy 46.91%\n",
            "Epoch 4220/15000 Cost: 0.709777 Accuracy 47.03%\n",
            "Epoch 4230/15000 Cost: 0.709700 Accuracy 47.03%\n",
            "Epoch 4240/15000 Cost: 0.709624 Accuracy 47.03%\n",
            "Epoch 4250/15000 Cost: 0.709547 Accuracy 47.03%\n",
            "Epoch 4260/15000 Cost: 0.709471 Accuracy 47.03%\n",
            "Epoch 4270/15000 Cost: 0.709395 Accuracy 47.03%\n",
            "Epoch 4280/15000 Cost: 0.709318 Accuracy 47.03%\n",
            "Epoch 4290/15000 Cost: 0.709242 Accuracy 47.03%\n",
            "Epoch 4300/15000 Cost: 0.709165 Accuracy 47.03%\n",
            "Epoch 4310/15000 Cost: 0.709089 Accuracy 47.25%\n",
            "Epoch 4320/15000 Cost: 0.709012 Accuracy 47.25%\n",
            "Epoch 4330/15000 Cost: 0.708936 Accuracy 47.36%\n",
            "Epoch 4340/15000 Cost: 0.708860 Accuracy 47.36%\n",
            "Epoch 4350/15000 Cost: 0.708784 Accuracy 47.36%\n",
            "Epoch 4360/15000 Cost: 0.708707 Accuracy 47.36%\n",
            "Epoch 4370/15000 Cost: 0.708631 Accuracy 47.36%\n",
            "Epoch 4380/15000 Cost: 0.708555 Accuracy 47.36%\n",
            "Epoch 4390/15000 Cost: 0.708479 Accuracy 47.36%\n",
            "Epoch 4400/15000 Cost: 0.708403 Accuracy 47.36%\n",
            "Epoch 4410/15000 Cost: 0.708326 Accuracy 47.36%\n",
            "Epoch 4420/15000 Cost: 0.708250 Accuracy 47.36%\n",
            "Epoch 4430/15000 Cost: 0.708174 Accuracy 47.36%\n",
            "Epoch 4440/15000 Cost: 0.708098 Accuracy 47.36%\n",
            "Epoch 4450/15000 Cost: 0.708022 Accuracy 47.36%\n",
            "Epoch 4460/15000 Cost: 0.707946 Accuracy 47.36%\n",
            "Epoch 4470/15000 Cost: 0.707870 Accuracy 47.59%\n",
            "Epoch 4480/15000 Cost: 0.707794 Accuracy 47.59%\n",
            "Epoch 4490/15000 Cost: 0.707718 Accuracy 47.92%\n",
            "Epoch 4500/15000 Cost: 0.707642 Accuracy 47.92%\n",
            "Epoch 4510/15000 Cost: 0.707566 Accuracy 48.04%\n",
            "Epoch 4520/15000 Cost: 0.707491 Accuracy 48.04%\n",
            "Epoch 4530/15000 Cost: 0.707415 Accuracy 48.04%\n",
            "Epoch 4540/15000 Cost: 0.707339 Accuracy 48.04%\n",
            "Epoch 4550/15000 Cost: 0.707263 Accuracy 48.04%\n",
            "Epoch 4560/15000 Cost: 0.707187 Accuracy 48.15%\n",
            "Epoch 4570/15000 Cost: 0.707112 Accuracy 48.15%\n",
            "Epoch 4580/15000 Cost: 0.707036 Accuracy 48.15%\n",
            "Epoch 4590/15000 Cost: 0.706960 Accuracy 48.15%\n",
            "Epoch 4600/15000 Cost: 0.706885 Accuracy 48.26%\n",
            "Epoch 4610/15000 Cost: 0.706809 Accuracy 48.26%\n",
            "Epoch 4620/15000 Cost: 0.706733 Accuracy 48.26%\n",
            "Epoch 4630/15000 Cost: 0.706658 Accuracy 48.26%\n",
            "Epoch 4640/15000 Cost: 0.706582 Accuracy 48.37%\n",
            "Epoch 4650/15000 Cost: 0.706507 Accuracy 48.37%\n",
            "Epoch 4660/15000 Cost: 0.706431 Accuracy 48.48%\n",
            "Epoch 4670/15000 Cost: 0.706356 Accuracy 48.48%\n",
            "Epoch 4680/15000 Cost: 0.706280 Accuracy 48.48%\n",
            "Epoch 4690/15000 Cost: 0.706205 Accuracy 48.48%\n",
            "Epoch 4700/15000 Cost: 0.706129 Accuracy 48.48%\n",
            "Epoch 4710/15000 Cost: 0.706054 Accuracy 48.48%\n",
            "Epoch 4720/15000 Cost: 0.705979 Accuracy 48.48%\n",
            "Epoch 4730/15000 Cost: 0.705903 Accuracy 48.48%\n",
            "Epoch 4740/15000 Cost: 0.705828 Accuracy 48.48%\n",
            "Epoch 4750/15000 Cost: 0.705753 Accuracy 48.48%\n",
            "Epoch 4760/15000 Cost: 0.705678 Accuracy 48.48%\n",
            "Epoch 4770/15000 Cost: 0.705602 Accuracy 48.48%\n",
            "Epoch 4780/15000 Cost: 0.705527 Accuracy 48.48%\n",
            "Epoch 4790/15000 Cost: 0.705452 Accuracy 48.48%\n",
            "Epoch 4800/15000 Cost: 0.705377 Accuracy 48.48%\n",
            "Epoch 4810/15000 Cost: 0.705302 Accuracy 48.48%\n",
            "Epoch 4820/15000 Cost: 0.705226 Accuracy 48.48%\n",
            "Epoch 4830/15000 Cost: 0.705151 Accuracy 48.48%\n",
            "Epoch 4840/15000 Cost: 0.705076 Accuracy 48.48%\n",
            "Epoch 4850/15000 Cost: 0.705001 Accuracy 48.48%\n",
            "Epoch 4860/15000 Cost: 0.704926 Accuracy 48.48%\n",
            "Epoch 4870/15000 Cost: 0.704851 Accuracy 48.48%\n",
            "Epoch 4880/15000 Cost: 0.704776 Accuracy 48.48%\n",
            "Epoch 4890/15000 Cost: 0.704701 Accuracy 48.48%\n",
            "Epoch 4900/15000 Cost: 0.704626 Accuracy 48.48%\n",
            "Epoch 4910/15000 Cost: 0.704551 Accuracy 48.60%\n",
            "Epoch 4920/15000 Cost: 0.704477 Accuracy 48.71%\n",
            "Epoch 4930/15000 Cost: 0.704402 Accuracy 48.71%\n",
            "Epoch 4940/15000 Cost: 0.704327 Accuracy 48.71%\n",
            "Epoch 4950/15000 Cost: 0.704252 Accuracy 48.71%\n",
            "Epoch 4960/15000 Cost: 0.704178 Accuracy 48.82%\n",
            "Epoch 4970/15000 Cost: 0.704103 Accuracy 48.82%\n",
            "Epoch 4980/15000 Cost: 0.704028 Accuracy 49.49%\n",
            "Epoch 4990/15000 Cost: 0.703953 Accuracy 49.49%\n",
            "Epoch 5000/15000 Cost: 0.703879 Accuracy 49.49%\n",
            "Epoch 5010/15000 Cost: 0.703804 Accuracy 49.61%\n",
            "Epoch 5020/15000 Cost: 0.703729 Accuracy 49.61%\n",
            "Epoch 5030/15000 Cost: 0.703655 Accuracy 49.61%\n",
            "Epoch 5040/15000 Cost: 0.703580 Accuracy 49.61%\n",
            "Epoch 5050/15000 Cost: 0.703506 Accuracy 49.61%\n",
            "Epoch 5060/15000 Cost: 0.703431 Accuracy 49.72%\n",
            "Epoch 5070/15000 Cost: 0.703357 Accuracy 49.72%\n",
            "Epoch 5080/15000 Cost: 0.703282 Accuracy 49.72%\n",
            "Epoch 5090/15000 Cost: 0.703208 Accuracy 49.72%\n",
            "Epoch 5100/15000 Cost: 0.703133 Accuracy 49.72%\n",
            "Epoch 5110/15000 Cost: 0.703059 Accuracy 49.83%\n",
            "Epoch 5120/15000 Cost: 0.702984 Accuracy 49.83%\n",
            "Epoch 5130/15000 Cost: 0.702910 Accuracy 49.72%\n",
            "Epoch 5140/15000 Cost: 0.702836 Accuracy 49.72%\n",
            "Epoch 5150/15000 Cost: 0.702761 Accuracy 49.72%\n",
            "Epoch 5160/15000 Cost: 0.702687 Accuracy 49.72%\n",
            "Epoch 5170/15000 Cost: 0.702613 Accuracy 49.61%\n",
            "Epoch 5180/15000 Cost: 0.702539 Accuracy 49.61%\n",
            "Epoch 5190/15000 Cost: 0.702464 Accuracy 49.61%\n",
            "Epoch 5200/15000 Cost: 0.702390 Accuracy 49.72%\n",
            "Epoch 5210/15000 Cost: 0.702316 Accuracy 49.72%\n",
            "Epoch 5220/15000 Cost: 0.702242 Accuracy 49.72%\n",
            "Epoch 5230/15000 Cost: 0.702168 Accuracy 49.72%\n",
            "Epoch 5240/15000 Cost: 0.702094 Accuracy 49.72%\n",
            "Epoch 5250/15000 Cost: 0.702020 Accuracy 49.83%\n",
            "Epoch 5260/15000 Cost: 0.701945 Accuracy 49.83%\n",
            "Epoch 5270/15000 Cost: 0.701871 Accuracy 49.94%\n",
            "Epoch 5280/15000 Cost: 0.701797 Accuracy 49.94%\n",
            "Epoch 5290/15000 Cost: 0.701723 Accuracy 50.06%\n",
            "Epoch 5300/15000 Cost: 0.701649 Accuracy 50.06%\n",
            "Epoch 5310/15000 Cost: 0.701576 Accuracy 50.06%\n",
            "Epoch 5320/15000 Cost: 0.701502 Accuracy 50.06%\n",
            "Epoch 5330/15000 Cost: 0.701428 Accuracy 50.17%\n",
            "Epoch 5340/15000 Cost: 0.701354 Accuracy 50.17%\n",
            "Epoch 5350/15000 Cost: 0.701280 Accuracy 50.17%\n",
            "Epoch 5360/15000 Cost: 0.701206 Accuracy 50.17%\n",
            "Epoch 5370/15000 Cost: 0.701132 Accuracy 50.17%\n",
            "Epoch 5380/15000 Cost: 0.701059 Accuracy 50.17%\n",
            "Epoch 5390/15000 Cost: 0.700985 Accuracy 50.17%\n",
            "Epoch 5400/15000 Cost: 0.700911 Accuracy 50.17%\n",
            "Epoch 5410/15000 Cost: 0.700838 Accuracy 50.17%\n",
            "Epoch 5420/15000 Cost: 0.700764 Accuracy 50.17%\n",
            "Epoch 5430/15000 Cost: 0.700690 Accuracy 50.17%\n",
            "Epoch 5440/15000 Cost: 0.700617 Accuracy 50.17%\n",
            "Epoch 5450/15000 Cost: 0.700543 Accuracy 50.17%\n",
            "Epoch 5460/15000 Cost: 0.700470 Accuracy 50.28%\n",
            "Epoch 5470/15000 Cost: 0.700396 Accuracy 50.28%\n",
            "Epoch 5480/15000 Cost: 0.700323 Accuracy 50.28%\n",
            "Epoch 5490/15000 Cost: 0.700249 Accuracy 50.28%\n",
            "Epoch 5500/15000 Cost: 0.700176 Accuracy 50.28%\n",
            "Epoch 5510/15000 Cost: 0.700102 Accuracy 50.28%\n",
            "Epoch 5520/15000 Cost: 0.700029 Accuracy 50.28%\n",
            "Epoch 5530/15000 Cost: 0.699955 Accuracy 50.28%\n",
            "Epoch 5540/15000 Cost: 0.699882 Accuracy 50.28%\n",
            "Epoch 5550/15000 Cost: 0.699809 Accuracy 50.28%\n",
            "Epoch 5560/15000 Cost: 0.699735 Accuracy 50.39%\n",
            "Epoch 5570/15000 Cost: 0.699662 Accuracy 50.39%\n",
            "Epoch 5580/15000 Cost: 0.699589 Accuracy 50.39%\n",
            "Epoch 5590/15000 Cost: 0.699516 Accuracy 50.39%\n",
            "Epoch 5600/15000 Cost: 0.699442 Accuracy 50.39%\n",
            "Epoch 5610/15000 Cost: 0.699369 Accuracy 50.39%\n",
            "Epoch 5620/15000 Cost: 0.699296 Accuracy 50.39%\n",
            "Epoch 5630/15000 Cost: 0.699223 Accuracy 50.39%\n",
            "Epoch 5640/15000 Cost: 0.699150 Accuracy 50.39%\n",
            "Epoch 5650/15000 Cost: 0.699077 Accuracy 50.51%\n",
            "Epoch 5660/15000 Cost: 0.699004 Accuracy 50.51%\n",
            "Epoch 5670/15000 Cost: 0.698931 Accuracy 50.51%\n",
            "Epoch 5680/15000 Cost: 0.698857 Accuracy 50.51%\n",
            "Epoch 5690/15000 Cost: 0.698785 Accuracy 50.51%\n",
            "Epoch 5700/15000 Cost: 0.698712 Accuracy 50.51%\n",
            "Epoch 5710/15000 Cost: 0.698639 Accuracy 50.51%\n",
            "Epoch 5720/15000 Cost: 0.698566 Accuracy 50.51%\n",
            "Epoch 5730/15000 Cost: 0.698493 Accuracy 50.51%\n",
            "Epoch 5740/15000 Cost: 0.698420 Accuracy 50.51%\n",
            "Epoch 5750/15000 Cost: 0.698347 Accuracy 50.51%\n",
            "Epoch 5760/15000 Cost: 0.698274 Accuracy 50.51%\n",
            "Epoch 5770/15000 Cost: 0.698201 Accuracy 50.51%\n",
            "Epoch 5780/15000 Cost: 0.698129 Accuracy 50.51%\n",
            "Epoch 5790/15000 Cost: 0.698056 Accuracy 50.51%\n",
            "Epoch 5800/15000 Cost: 0.697983 Accuracy 50.51%\n",
            "Epoch 5810/15000 Cost: 0.697910 Accuracy 50.51%\n",
            "Epoch 5820/15000 Cost: 0.697838 Accuracy 50.51%\n",
            "Epoch 5830/15000 Cost: 0.697765 Accuracy 50.51%\n",
            "Epoch 5840/15000 Cost: 0.697692 Accuracy 50.51%\n",
            "Epoch 5850/15000 Cost: 0.697620 Accuracy 50.51%\n",
            "Epoch 5860/15000 Cost: 0.697547 Accuracy 50.51%\n",
            "Epoch 5870/15000 Cost: 0.697475 Accuracy 50.51%\n",
            "Epoch 5880/15000 Cost: 0.697402 Accuracy 50.62%\n",
            "Epoch 5890/15000 Cost: 0.697330 Accuracy 50.62%\n",
            "Epoch 5900/15000 Cost: 0.697257 Accuracy 50.73%\n",
            "Epoch 5910/15000 Cost: 0.697185 Accuracy 50.84%\n",
            "Epoch 5920/15000 Cost: 0.697112 Accuracy 50.84%\n",
            "Epoch 5930/15000 Cost: 0.697040 Accuracy 50.84%\n",
            "Epoch 5940/15000 Cost: 0.696967 Accuracy 50.84%\n",
            "Epoch 5950/15000 Cost: 0.696895 Accuracy 50.84%\n",
            "Epoch 5960/15000 Cost: 0.696823 Accuracy 50.84%\n",
            "Epoch 5970/15000 Cost: 0.696750 Accuracy 50.84%\n",
            "Epoch 5980/15000 Cost: 0.696678 Accuracy 50.84%\n",
            "Epoch 5990/15000 Cost: 0.696606 Accuracy 50.84%\n",
            "Epoch 6000/15000 Cost: 0.696534 Accuracy 50.84%\n",
            "Epoch 6010/15000 Cost: 0.696461 Accuracy 50.84%\n",
            "Epoch 6020/15000 Cost: 0.696389 Accuracy 50.84%\n",
            "Epoch 6030/15000 Cost: 0.696317 Accuracy 50.62%\n",
            "Epoch 6040/15000 Cost: 0.696245 Accuracy 50.62%\n",
            "Epoch 6050/15000 Cost: 0.696172 Accuracy 50.62%\n",
            "Epoch 6060/15000 Cost: 0.696100 Accuracy 50.62%\n",
            "Epoch 6070/15000 Cost: 0.696028 Accuracy 50.62%\n",
            "Epoch 6080/15000 Cost: 0.695956 Accuracy 50.62%\n",
            "Epoch 6090/15000 Cost: 0.695884 Accuracy 50.62%\n",
            "Epoch 6100/15000 Cost: 0.695812 Accuracy 50.62%\n",
            "Epoch 6110/15000 Cost: 0.695740 Accuracy 50.62%\n",
            "Epoch 6120/15000 Cost: 0.695668 Accuracy 50.62%\n",
            "Epoch 6130/15000 Cost: 0.695596 Accuracy 50.62%\n",
            "Epoch 6140/15000 Cost: 0.695524 Accuracy 50.73%\n",
            "Epoch 6150/15000 Cost: 0.695452 Accuracy 50.73%\n",
            "Epoch 6160/15000 Cost: 0.695380 Accuracy 50.73%\n",
            "Epoch 6170/15000 Cost: 0.695309 Accuracy 50.84%\n",
            "Epoch 6180/15000 Cost: 0.695237 Accuracy 50.84%\n",
            "Epoch 6190/15000 Cost: 0.695165 Accuracy 50.84%\n",
            "Epoch 6200/15000 Cost: 0.695093 Accuracy 50.95%\n",
            "Epoch 6210/15000 Cost: 0.695021 Accuracy 50.95%\n",
            "Epoch 6220/15000 Cost: 0.694950 Accuracy 50.95%\n",
            "Epoch 6230/15000 Cost: 0.694878 Accuracy 50.95%\n",
            "Epoch 6240/15000 Cost: 0.694806 Accuracy 50.95%\n",
            "Epoch 6250/15000 Cost: 0.694735 Accuracy 50.95%\n",
            "Epoch 6260/15000 Cost: 0.694663 Accuracy 50.95%\n",
            "Epoch 6270/15000 Cost: 0.694591 Accuracy 50.95%\n",
            "Epoch 6280/15000 Cost: 0.694520 Accuracy 50.95%\n",
            "Epoch 6290/15000 Cost: 0.694448 Accuracy 50.95%\n",
            "Epoch 6300/15000 Cost: 0.694377 Accuracy 50.95%\n",
            "Epoch 6310/15000 Cost: 0.694305 Accuracy 50.95%\n",
            "Epoch 6320/15000 Cost: 0.694234 Accuracy 50.95%\n",
            "Epoch 6330/15000 Cost: 0.694162 Accuracy 50.95%\n",
            "Epoch 6340/15000 Cost: 0.694091 Accuracy 51.07%\n",
            "Epoch 6350/15000 Cost: 0.694019 Accuracy 51.07%\n",
            "Epoch 6360/15000 Cost: 0.693948 Accuracy 51.29%\n",
            "Epoch 6370/15000 Cost: 0.693877 Accuracy 51.29%\n",
            "Epoch 6380/15000 Cost: 0.693805 Accuracy 51.63%\n",
            "Epoch 6390/15000 Cost: 0.693734 Accuracy 51.63%\n",
            "Epoch 6400/15000 Cost: 0.693663 Accuracy 51.63%\n",
            "Epoch 6410/15000 Cost: 0.693591 Accuracy 51.63%\n",
            "Epoch 6420/15000 Cost: 0.693520 Accuracy 51.74%\n",
            "Epoch 6430/15000 Cost: 0.693449 Accuracy 51.74%\n",
            "Epoch 6440/15000 Cost: 0.693378 Accuracy 51.74%\n",
            "Epoch 6450/15000 Cost: 0.693306 Accuracy 51.74%\n",
            "Epoch 6460/15000 Cost: 0.693235 Accuracy 51.85%\n",
            "Epoch 6470/15000 Cost: 0.693164 Accuracy 51.85%\n",
            "Epoch 6480/15000 Cost: 0.693093 Accuracy 51.96%\n",
            "Epoch 6490/15000 Cost: 0.693022 Accuracy 51.96%\n",
            "Epoch 6500/15000 Cost: 0.692951 Accuracy 52.19%\n",
            "Epoch 6510/15000 Cost: 0.692880 Accuracy 52.19%\n",
            "Epoch 6520/15000 Cost: 0.692809 Accuracy 52.30%\n",
            "Epoch 6530/15000 Cost: 0.692738 Accuracy 52.41%\n",
            "Epoch 6540/15000 Cost: 0.692667 Accuracy 52.53%\n",
            "Epoch 6550/15000 Cost: 0.692596 Accuracy 52.53%\n",
            "Epoch 6560/15000 Cost: 0.692525 Accuracy 52.53%\n",
            "Epoch 6570/15000 Cost: 0.692454 Accuracy 52.53%\n",
            "Epoch 6580/15000 Cost: 0.692383 Accuracy 52.53%\n",
            "Epoch 6590/15000 Cost: 0.692312 Accuracy 52.53%\n",
            "Epoch 6600/15000 Cost: 0.692241 Accuracy 52.53%\n",
            "Epoch 6610/15000 Cost: 0.692171 Accuracy 52.53%\n",
            "Epoch 6620/15000 Cost: 0.692100 Accuracy 52.53%\n",
            "Epoch 6630/15000 Cost: 0.692029 Accuracy 52.53%\n",
            "Epoch 6640/15000 Cost: 0.691958 Accuracy 52.53%\n",
            "Epoch 6650/15000 Cost: 0.691888 Accuracy 52.53%\n",
            "Epoch 6660/15000 Cost: 0.691817 Accuracy 52.53%\n",
            "Epoch 6670/15000 Cost: 0.691746 Accuracy 52.64%\n",
            "Epoch 6680/15000 Cost: 0.691676 Accuracy 52.64%\n",
            "Epoch 6690/15000 Cost: 0.691605 Accuracy 52.64%\n",
            "Epoch 6700/15000 Cost: 0.691535 Accuracy 52.64%\n",
            "Epoch 6710/15000 Cost: 0.691464 Accuracy 52.64%\n",
            "Epoch 6720/15000 Cost: 0.691394 Accuracy 52.64%\n",
            "Epoch 6730/15000 Cost: 0.691323 Accuracy 52.64%\n",
            "Epoch 6740/15000 Cost: 0.691253 Accuracy 52.75%\n",
            "Epoch 6750/15000 Cost: 0.691182 Accuracy 52.86%\n",
            "Epoch 6760/15000 Cost: 0.691112 Accuracy 52.86%\n",
            "Epoch 6770/15000 Cost: 0.691041 Accuracy 52.97%\n",
            "Epoch 6780/15000 Cost: 0.690971 Accuracy 52.97%\n",
            "Epoch 6790/15000 Cost: 0.690901 Accuracy 52.97%\n",
            "Epoch 6800/15000 Cost: 0.690830 Accuracy 52.97%\n",
            "Epoch 6810/15000 Cost: 0.690760 Accuracy 52.97%\n",
            "Epoch 6820/15000 Cost: 0.690690 Accuracy 52.97%\n",
            "Epoch 6830/15000 Cost: 0.690619 Accuracy 53.09%\n",
            "Epoch 6840/15000 Cost: 0.690549 Accuracy 53.09%\n",
            "Epoch 6850/15000 Cost: 0.690479 Accuracy 53.09%\n",
            "Epoch 6860/15000 Cost: 0.690409 Accuracy 53.09%\n",
            "Epoch 6870/15000 Cost: 0.690338 Accuracy 53.09%\n",
            "Epoch 6880/15000 Cost: 0.690268 Accuracy 53.09%\n",
            "Epoch 6890/15000 Cost: 0.690198 Accuracy 53.09%\n",
            "Epoch 6900/15000 Cost: 0.690128 Accuracy 53.09%\n",
            "Epoch 6910/15000 Cost: 0.690058 Accuracy 53.09%\n",
            "Epoch 6920/15000 Cost: 0.689988 Accuracy 53.09%\n",
            "Epoch 6930/15000 Cost: 0.689918 Accuracy 53.20%\n",
            "Epoch 6940/15000 Cost: 0.689848 Accuracy 53.31%\n",
            "Epoch 6950/15000 Cost: 0.689778 Accuracy 53.31%\n",
            "Epoch 6960/15000 Cost: 0.689708 Accuracy 53.42%\n",
            "Epoch 6970/15000 Cost: 0.689638 Accuracy 53.42%\n",
            "Epoch 6980/15000 Cost: 0.689568 Accuracy 53.42%\n",
            "Epoch 6990/15000 Cost: 0.689498 Accuracy 53.42%\n",
            "Epoch 7000/15000 Cost: 0.689428 Accuracy 53.42%\n",
            "Epoch 7010/15000 Cost: 0.689358 Accuracy 53.42%\n",
            "Epoch 7020/15000 Cost: 0.689288 Accuracy 53.42%\n",
            "Epoch 7030/15000 Cost: 0.689219 Accuracy 53.42%\n",
            "Epoch 7040/15000 Cost: 0.689149 Accuracy 53.42%\n",
            "Epoch 7050/15000 Cost: 0.689079 Accuracy 53.54%\n",
            "Epoch 7060/15000 Cost: 0.689009 Accuracy 53.54%\n",
            "Epoch 7070/15000 Cost: 0.688940 Accuracy 53.65%\n",
            "Epoch 7080/15000 Cost: 0.688870 Accuracy 53.65%\n",
            "Epoch 7090/15000 Cost: 0.688800 Accuracy 53.87%\n",
            "Epoch 7100/15000 Cost: 0.688731 Accuracy 53.87%\n",
            "Epoch 7110/15000 Cost: 0.688661 Accuracy 54.10%\n",
            "Epoch 7120/15000 Cost: 0.688592 Accuracy 54.10%\n",
            "Epoch 7130/15000 Cost: 0.688522 Accuracy 54.21%\n",
            "Epoch 7140/15000 Cost: 0.688452 Accuracy 54.32%\n",
            "Epoch 7150/15000 Cost: 0.688383 Accuracy 54.32%\n",
            "Epoch 7160/15000 Cost: 0.688313 Accuracy 54.32%\n",
            "Epoch 7170/15000 Cost: 0.688244 Accuracy 54.32%\n",
            "Epoch 7180/15000 Cost: 0.688175 Accuracy 54.32%\n",
            "Epoch 7190/15000 Cost: 0.688105 Accuracy 54.32%\n",
            "Epoch 7200/15000 Cost: 0.688036 Accuracy 54.43%\n",
            "Epoch 7210/15000 Cost: 0.687966 Accuracy 54.43%\n",
            "Epoch 7220/15000 Cost: 0.687897 Accuracy 54.55%\n",
            "Epoch 7230/15000 Cost: 0.687828 Accuracy 54.66%\n",
            "Epoch 7240/15000 Cost: 0.687758 Accuracy 54.66%\n",
            "Epoch 7250/15000 Cost: 0.687689 Accuracy 54.88%\n",
            "Epoch 7260/15000 Cost: 0.687620 Accuracy 54.88%\n",
            "Epoch 7270/15000 Cost: 0.687551 Accuracy 54.88%\n",
            "Epoch 7280/15000 Cost: 0.687481 Accuracy 54.88%\n",
            "Epoch 7290/15000 Cost: 0.687412 Accuracy 54.88%\n",
            "Epoch 7300/15000 Cost: 0.687343 Accuracy 54.88%\n",
            "Epoch 7310/15000 Cost: 0.687274 Accuracy 54.88%\n",
            "Epoch 7320/15000 Cost: 0.687205 Accuracy 54.88%\n",
            "Epoch 7330/15000 Cost: 0.687136 Accuracy 54.88%\n",
            "Epoch 7340/15000 Cost: 0.687066 Accuracy 54.99%\n",
            "Epoch 7350/15000 Cost: 0.686997 Accuracy 54.99%\n",
            "Epoch 7360/15000 Cost: 0.686929 Accuracy 54.99%\n",
            "Epoch 7370/15000 Cost: 0.686859 Accuracy 54.99%\n",
            "Epoch 7380/15000 Cost: 0.686790 Accuracy 54.99%\n",
            "Epoch 7390/15000 Cost: 0.686722 Accuracy 54.99%\n",
            "Epoch 7400/15000 Cost: 0.686653 Accuracy 54.99%\n",
            "Epoch 7410/15000 Cost: 0.686584 Accuracy 54.99%\n",
            "Epoch 7420/15000 Cost: 0.686515 Accuracy 55.11%\n",
            "Epoch 7430/15000 Cost: 0.686446 Accuracy 55.11%\n",
            "Epoch 7440/15000 Cost: 0.686377 Accuracy 55.11%\n",
            "Epoch 7450/15000 Cost: 0.686308 Accuracy 55.33%\n",
            "Epoch 7460/15000 Cost: 0.686240 Accuracy 55.33%\n",
            "Epoch 7470/15000 Cost: 0.686171 Accuracy 55.33%\n",
            "Epoch 7480/15000 Cost: 0.686102 Accuracy 55.33%\n",
            "Epoch 7490/15000 Cost: 0.686033 Accuracy 55.33%\n",
            "Epoch 7500/15000 Cost: 0.685965 Accuracy 55.67%\n",
            "Epoch 7510/15000 Cost: 0.685896 Accuracy 55.67%\n",
            "Epoch 7520/15000 Cost: 0.685828 Accuracy 55.67%\n",
            "Epoch 7530/15000 Cost: 0.685759 Accuracy 55.67%\n",
            "Epoch 7540/15000 Cost: 0.685690 Accuracy 55.67%\n",
            "Epoch 7550/15000 Cost: 0.685622 Accuracy 55.67%\n",
            "Epoch 7560/15000 Cost: 0.685553 Accuracy 55.78%\n",
            "Epoch 7570/15000 Cost: 0.685485 Accuracy 55.78%\n",
            "Epoch 7580/15000 Cost: 0.685416 Accuracy 55.78%\n",
            "Epoch 7590/15000 Cost: 0.685348 Accuracy 55.78%\n",
            "Epoch 7600/15000 Cost: 0.685279 Accuracy 55.89%\n",
            "Epoch 7610/15000 Cost: 0.685211 Accuracy 55.89%\n",
            "Epoch 7620/15000 Cost: 0.685143 Accuracy 55.89%\n",
            "Epoch 7630/15000 Cost: 0.685074 Accuracy 56.12%\n",
            "Epoch 7640/15000 Cost: 0.685006 Accuracy 57.69%\n",
            "Epoch 7650/15000 Cost: 0.684938 Accuracy 57.69%\n",
            "Epoch 7660/15000 Cost: 0.684869 Accuracy 57.69%\n",
            "Epoch 7670/15000 Cost: 0.684801 Accuracy 59.15%\n",
            "Epoch 7680/15000 Cost: 0.684733 Accuracy 59.15%\n",
            "Epoch 7690/15000 Cost: 0.684665 Accuracy 59.03%\n",
            "Epoch 7700/15000 Cost: 0.684596 Accuracy 59.03%\n",
            "Epoch 7710/15000 Cost: 0.684528 Accuracy 59.03%\n",
            "Epoch 7720/15000 Cost: 0.684460 Accuracy 59.03%\n",
            "Epoch 7730/15000 Cost: 0.684392 Accuracy 59.03%\n",
            "Epoch 7740/15000 Cost: 0.684324 Accuracy 59.03%\n",
            "Epoch 7750/15000 Cost: 0.684256 Accuracy 59.03%\n",
            "Epoch 7760/15000 Cost: 0.684188 Accuracy 59.03%\n",
            "Epoch 7770/15000 Cost: 0.684120 Accuracy 59.03%\n",
            "Epoch 7780/15000 Cost: 0.684052 Accuracy 59.15%\n",
            "Epoch 7790/15000 Cost: 0.683984 Accuracy 59.26%\n",
            "Epoch 7800/15000 Cost: 0.683916 Accuracy 59.26%\n",
            "Epoch 7810/15000 Cost: 0.683848 Accuracy 59.48%\n",
            "Epoch 7820/15000 Cost: 0.683780 Accuracy 59.48%\n",
            "Epoch 7830/15000 Cost: 0.683712 Accuracy 59.48%\n",
            "Epoch 7840/15000 Cost: 0.683644 Accuracy 59.48%\n",
            "Epoch 7850/15000 Cost: 0.683576 Accuracy 59.60%\n",
            "Epoch 7860/15000 Cost: 0.683508 Accuracy 59.71%\n",
            "Epoch 7870/15000 Cost: 0.683441 Accuracy 59.71%\n",
            "Epoch 7880/15000 Cost: 0.683373 Accuracy 59.71%\n",
            "Epoch 7890/15000 Cost: 0.683305 Accuracy 59.71%\n",
            "Epoch 7900/15000 Cost: 0.683237 Accuracy 59.71%\n",
            "Epoch 7910/15000 Cost: 0.683170 Accuracy 59.71%\n",
            "Epoch 7920/15000 Cost: 0.683102 Accuracy 59.60%\n",
            "Epoch 7930/15000 Cost: 0.683034 Accuracy 59.71%\n",
            "Epoch 7940/15000 Cost: 0.682967 Accuracy 59.71%\n",
            "Epoch 7950/15000 Cost: 0.682899 Accuracy 59.71%\n",
            "Epoch 7960/15000 Cost: 0.682831 Accuracy 59.71%\n",
            "Epoch 7970/15000 Cost: 0.682764 Accuracy 59.71%\n",
            "Epoch 7980/15000 Cost: 0.682696 Accuracy 59.71%\n",
            "Epoch 7990/15000 Cost: 0.682629 Accuracy 59.93%\n",
            "Epoch 8000/15000 Cost: 0.682562 Accuracy 59.93%\n",
            "Epoch 8010/15000 Cost: 0.682494 Accuracy 59.93%\n",
            "Epoch 8020/15000 Cost: 0.682426 Accuracy 59.93%\n",
            "Epoch 8030/15000 Cost: 0.682359 Accuracy 59.93%\n",
            "Epoch 8040/15000 Cost: 0.682292 Accuracy 59.93%\n",
            "Epoch 8050/15000 Cost: 0.682224 Accuracy 59.93%\n",
            "Epoch 8060/15000 Cost: 0.682157 Accuracy 59.93%\n",
            "Epoch 8070/15000 Cost: 0.682090 Accuracy 60.04%\n",
            "Epoch 8080/15000 Cost: 0.682022 Accuracy 60.04%\n",
            "Epoch 8090/15000 Cost: 0.681955 Accuracy 60.04%\n",
            "Epoch 8100/15000 Cost: 0.681888 Accuracy 60.04%\n",
            "Epoch 8110/15000 Cost: 0.681821 Accuracy 60.04%\n",
            "Epoch 8120/15000 Cost: 0.681753 Accuracy 60.04%\n",
            "Epoch 8130/15000 Cost: 0.681686 Accuracy 60.27%\n",
            "Epoch 8140/15000 Cost: 0.681619 Accuracy 60.27%\n",
            "Epoch 8150/15000 Cost: 0.681552 Accuracy 60.38%\n",
            "Epoch 8160/15000 Cost: 0.681485 Accuracy 60.38%\n",
            "Epoch 8170/15000 Cost: 0.681418 Accuracy 60.38%\n",
            "Epoch 8180/15000 Cost: 0.681351 Accuracy 60.72%\n",
            "Epoch 8190/15000 Cost: 0.681283 Accuracy 60.72%\n",
            "Epoch 8200/15000 Cost: 0.681216 Accuracy 60.72%\n",
            "Epoch 8210/15000 Cost: 0.681149 Accuracy 60.72%\n",
            "Epoch 8220/15000 Cost: 0.681082 Accuracy 60.72%\n",
            "Epoch 8230/15000 Cost: 0.681016 Accuracy 60.72%\n",
            "Epoch 8240/15000 Cost: 0.680948 Accuracy 60.83%\n",
            "Epoch 8250/15000 Cost: 0.680882 Accuracy 60.83%\n",
            "Epoch 8260/15000 Cost: 0.680815 Accuracy 60.83%\n",
            "Epoch 8270/15000 Cost: 0.680748 Accuracy 60.83%\n",
            "Epoch 8280/15000 Cost: 0.680681 Accuracy 60.83%\n",
            "Epoch 8290/15000 Cost: 0.680614 Accuracy 60.83%\n",
            "Epoch 8300/15000 Cost: 0.680547 Accuracy 60.83%\n",
            "Epoch 8310/15000 Cost: 0.680481 Accuracy 60.83%\n",
            "Epoch 8320/15000 Cost: 0.680414 Accuracy 60.83%\n",
            "Epoch 8330/15000 Cost: 0.680347 Accuracy 60.83%\n",
            "Epoch 8340/15000 Cost: 0.680281 Accuracy 60.83%\n",
            "Epoch 8350/15000 Cost: 0.680214 Accuracy 60.83%\n",
            "Epoch 8360/15000 Cost: 0.680147 Accuracy 60.83%\n",
            "Epoch 8370/15000 Cost: 0.680081 Accuracy 60.83%\n",
            "Epoch 8380/15000 Cost: 0.680014 Accuracy 60.83%\n",
            "Epoch 8390/15000 Cost: 0.679947 Accuracy 60.83%\n",
            "Epoch 8400/15000 Cost: 0.679881 Accuracy 60.83%\n",
            "Epoch 8410/15000 Cost: 0.679814 Accuracy 60.83%\n",
            "Epoch 8420/15000 Cost: 0.679748 Accuracy 60.83%\n",
            "Epoch 8430/15000 Cost: 0.679681 Accuracy 60.94%\n",
            "Epoch 8440/15000 Cost: 0.679615 Accuracy 60.94%\n",
            "Epoch 8450/15000 Cost: 0.679548 Accuracy 60.94%\n",
            "Epoch 8460/15000 Cost: 0.679482 Accuracy 60.94%\n",
            "Epoch 8470/15000 Cost: 0.679416 Accuracy 60.94%\n",
            "Epoch 8480/15000 Cost: 0.679349 Accuracy 60.94%\n",
            "Epoch 8490/15000 Cost: 0.679283 Accuracy 60.94%\n",
            "Epoch 8500/15000 Cost: 0.679216 Accuracy 60.94%\n",
            "Epoch 8510/15000 Cost: 0.679150 Accuracy 60.94%\n",
            "Epoch 8520/15000 Cost: 0.679084 Accuracy 60.94%\n",
            "Epoch 8530/15000 Cost: 0.679018 Accuracy 60.94%\n",
            "Epoch 8540/15000 Cost: 0.678951 Accuracy 60.94%\n",
            "Epoch 8550/15000 Cost: 0.678885 Accuracy 61.05%\n",
            "Epoch 8560/15000 Cost: 0.678819 Accuracy 61.05%\n",
            "Epoch 8570/15000 Cost: 0.678753 Accuracy 61.05%\n",
            "Epoch 8580/15000 Cost: 0.678687 Accuracy 61.17%\n",
            "Epoch 8590/15000 Cost: 0.678621 Accuracy 61.17%\n",
            "Epoch 8600/15000 Cost: 0.678554 Accuracy 61.17%\n",
            "Epoch 8610/15000 Cost: 0.678488 Accuracy 61.17%\n",
            "Epoch 8620/15000 Cost: 0.678422 Accuracy 61.17%\n",
            "Epoch 8630/15000 Cost: 0.678356 Accuracy 61.17%\n",
            "Epoch 8640/15000 Cost: 0.678290 Accuracy 61.17%\n",
            "Epoch 8650/15000 Cost: 0.678224 Accuracy 61.17%\n",
            "Epoch 8660/15000 Cost: 0.678158 Accuracy 61.28%\n",
            "Epoch 8670/15000 Cost: 0.678092 Accuracy 61.50%\n",
            "Epoch 8680/15000 Cost: 0.678027 Accuracy 61.84%\n",
            "Epoch 8690/15000 Cost: 0.677961 Accuracy 61.84%\n",
            "Epoch 8700/15000 Cost: 0.677895 Accuracy 61.84%\n",
            "Epoch 8710/15000 Cost: 0.677829 Accuracy 61.84%\n",
            "Epoch 8720/15000 Cost: 0.677763 Accuracy 61.84%\n",
            "Epoch 8730/15000 Cost: 0.677697 Accuracy 61.95%\n",
            "Epoch 8740/15000 Cost: 0.677632 Accuracy 61.84%\n",
            "Epoch 8750/15000 Cost: 0.677566 Accuracy 61.84%\n",
            "Epoch 8760/15000 Cost: 0.677500 Accuracy 61.84%\n",
            "Epoch 8770/15000 Cost: 0.677434 Accuracy 61.84%\n",
            "Epoch 8780/15000 Cost: 0.677369 Accuracy 61.84%\n",
            "Epoch 8790/15000 Cost: 0.677303 Accuracy 61.95%\n",
            "Epoch 8800/15000 Cost: 0.677238 Accuracy 61.95%\n",
            "Epoch 8810/15000 Cost: 0.677172 Accuracy 61.95%\n",
            "Epoch 8820/15000 Cost: 0.677106 Accuracy 61.95%\n",
            "Epoch 8830/15000 Cost: 0.677041 Accuracy 61.95%\n",
            "Epoch 8840/15000 Cost: 0.676975 Accuracy 61.95%\n",
            "Epoch 8850/15000 Cost: 0.676910 Accuracy 61.95%\n",
            "Epoch 8860/15000 Cost: 0.676844 Accuracy 61.95%\n",
            "Epoch 8870/15000 Cost: 0.676779 Accuracy 61.95%\n",
            "Epoch 8880/15000 Cost: 0.676713 Accuracy 61.84%\n",
            "Epoch 8890/15000 Cost: 0.676648 Accuracy 61.84%\n",
            "Epoch 8900/15000 Cost: 0.676583 Accuracy 61.84%\n",
            "Epoch 8910/15000 Cost: 0.676517 Accuracy 61.84%\n",
            "Epoch 8920/15000 Cost: 0.676452 Accuracy 61.84%\n",
            "Epoch 8930/15000 Cost: 0.676387 Accuracy 61.84%\n",
            "Epoch 8940/15000 Cost: 0.676321 Accuracy 61.84%\n",
            "Epoch 8950/15000 Cost: 0.676256 Accuracy 61.84%\n",
            "Epoch 8960/15000 Cost: 0.676191 Accuracy 61.95%\n",
            "Epoch 8970/15000 Cost: 0.676126 Accuracy 61.95%\n",
            "Epoch 8980/15000 Cost: 0.676060 Accuracy 61.95%\n",
            "Epoch 8990/15000 Cost: 0.675995 Accuracy 61.95%\n",
            "Epoch 9000/15000 Cost: 0.675930 Accuracy 61.95%\n",
            "Epoch 9010/15000 Cost: 0.675865 Accuracy 61.95%\n",
            "Epoch 9020/15000 Cost: 0.675800 Accuracy 62.07%\n",
            "Epoch 9030/15000 Cost: 0.675735 Accuracy 62.07%\n",
            "Epoch 9040/15000 Cost: 0.675670 Accuracy 62.07%\n",
            "Epoch 9050/15000 Cost: 0.675605 Accuracy 62.18%\n",
            "Epoch 9060/15000 Cost: 0.675539 Accuracy 62.18%\n",
            "Epoch 9070/15000 Cost: 0.675475 Accuracy 62.18%\n",
            "Epoch 9080/15000 Cost: 0.675410 Accuracy 62.18%\n",
            "Epoch 9090/15000 Cost: 0.675345 Accuracy 62.18%\n",
            "Epoch 9100/15000 Cost: 0.675280 Accuracy 62.18%\n",
            "Epoch 9110/15000 Cost: 0.675215 Accuracy 62.18%\n",
            "Epoch 9120/15000 Cost: 0.675150 Accuracy 62.18%\n",
            "Epoch 9130/15000 Cost: 0.675085 Accuracy 62.18%\n",
            "Epoch 9140/15000 Cost: 0.675020 Accuracy 62.51%\n",
            "Epoch 9150/15000 Cost: 0.674955 Accuracy 62.51%\n",
            "Epoch 9160/15000 Cost: 0.674891 Accuracy 62.51%\n",
            "Epoch 9170/15000 Cost: 0.674826 Accuracy 62.51%\n",
            "Epoch 9180/15000 Cost: 0.674761 Accuracy 62.51%\n",
            "Epoch 9190/15000 Cost: 0.674697 Accuracy 62.85%\n",
            "Epoch 9200/15000 Cost: 0.674632 Accuracy 62.85%\n",
            "Epoch 9210/15000 Cost: 0.674567 Accuracy 62.85%\n",
            "Epoch 9220/15000 Cost: 0.674502 Accuracy 62.85%\n",
            "Epoch 9230/15000 Cost: 0.674438 Accuracy 62.85%\n",
            "Epoch 9240/15000 Cost: 0.674373 Accuracy 62.85%\n",
            "Epoch 9250/15000 Cost: 0.674309 Accuracy 62.85%\n",
            "Epoch 9260/15000 Cost: 0.674244 Accuracy 62.85%\n",
            "Epoch 9270/15000 Cost: 0.674180 Accuracy 62.85%\n",
            "Epoch 9280/15000 Cost: 0.674115 Accuracy 62.96%\n",
            "Epoch 9290/15000 Cost: 0.674051 Accuracy 62.96%\n",
            "Epoch 9300/15000 Cost: 0.673986 Accuracy 62.96%\n",
            "Epoch 9310/15000 Cost: 0.673922 Accuracy 62.96%\n",
            "Epoch 9320/15000 Cost: 0.673857 Accuracy 63.08%\n",
            "Epoch 9330/15000 Cost: 0.673793 Accuracy 63.08%\n",
            "Epoch 9340/15000 Cost: 0.673729 Accuracy 63.08%\n",
            "Epoch 9350/15000 Cost: 0.673664 Accuracy 63.08%\n",
            "Epoch 9360/15000 Cost: 0.673600 Accuracy 63.08%\n",
            "Epoch 9370/15000 Cost: 0.673535 Accuracy 63.08%\n",
            "Epoch 9380/15000 Cost: 0.673471 Accuracy 63.41%\n",
            "Epoch 9390/15000 Cost: 0.673407 Accuracy 63.41%\n",
            "Epoch 9400/15000 Cost: 0.673343 Accuracy 63.41%\n",
            "Epoch 9410/15000 Cost: 0.673279 Accuracy 63.41%\n",
            "Epoch 9420/15000 Cost: 0.673214 Accuracy 63.41%\n",
            "Epoch 9430/15000 Cost: 0.673150 Accuracy 63.41%\n",
            "Epoch 9440/15000 Cost: 0.673086 Accuracy 63.41%\n",
            "Epoch 9450/15000 Cost: 0.673022 Accuracy 63.41%\n",
            "Epoch 9460/15000 Cost: 0.672958 Accuracy 63.41%\n",
            "Epoch 9470/15000 Cost: 0.672894 Accuracy 63.41%\n",
            "Epoch 9480/15000 Cost: 0.672830 Accuracy 63.41%\n",
            "Epoch 9490/15000 Cost: 0.672766 Accuracy 63.41%\n",
            "Epoch 9500/15000 Cost: 0.672702 Accuracy 63.41%\n",
            "Epoch 9510/15000 Cost: 0.672638 Accuracy 63.41%\n",
            "Epoch 9520/15000 Cost: 0.672574 Accuracy 63.41%\n",
            "Epoch 9530/15000 Cost: 0.672510 Accuracy 63.41%\n",
            "Epoch 9540/15000 Cost: 0.672446 Accuracy 63.41%\n",
            "Epoch 9550/15000 Cost: 0.672382 Accuracy 63.41%\n",
            "Epoch 9560/15000 Cost: 0.672318 Accuracy 63.41%\n",
            "Epoch 9570/15000 Cost: 0.672254 Accuracy 63.41%\n",
            "Epoch 9580/15000 Cost: 0.672190 Accuracy 63.41%\n",
            "Epoch 9590/15000 Cost: 0.672127 Accuracy 63.41%\n",
            "Epoch 9600/15000 Cost: 0.672063 Accuracy 63.41%\n",
            "Epoch 9610/15000 Cost: 0.671999 Accuracy 63.41%\n",
            "Epoch 9620/15000 Cost: 0.671935 Accuracy 63.41%\n",
            "Epoch 9630/15000 Cost: 0.671872 Accuracy 63.41%\n",
            "Epoch 9640/15000 Cost: 0.671808 Accuracy 63.52%\n",
            "Epoch 9650/15000 Cost: 0.671744 Accuracy 63.52%\n",
            "Epoch 9660/15000 Cost: 0.671681 Accuracy 63.52%\n",
            "Epoch 9670/15000 Cost: 0.671617 Accuracy 63.64%\n",
            "Epoch 9680/15000 Cost: 0.671553 Accuracy 63.75%\n",
            "Epoch 9690/15000 Cost: 0.671490 Accuracy 63.75%\n",
            "Epoch 9700/15000 Cost: 0.671426 Accuracy 63.75%\n",
            "Epoch 9710/15000 Cost: 0.671363 Accuracy 63.75%\n",
            "Epoch 9720/15000 Cost: 0.671299 Accuracy 63.86%\n",
            "Epoch 9730/15000 Cost: 0.671236 Accuracy 63.86%\n",
            "Epoch 9740/15000 Cost: 0.671172 Accuracy 63.86%\n",
            "Epoch 9750/15000 Cost: 0.671109 Accuracy 63.86%\n",
            "Epoch 9760/15000 Cost: 0.671046 Accuracy 63.86%\n",
            "Epoch 9770/15000 Cost: 0.670982 Accuracy 63.86%\n",
            "Epoch 9780/15000 Cost: 0.670919 Accuracy 64.09%\n",
            "Epoch 9790/15000 Cost: 0.670856 Accuracy 64.09%\n",
            "Epoch 9800/15000 Cost: 0.670792 Accuracy 64.09%\n",
            "Epoch 9810/15000 Cost: 0.670729 Accuracy 64.09%\n",
            "Epoch 9820/15000 Cost: 0.670666 Accuracy 64.09%\n",
            "Epoch 9830/15000 Cost: 0.670603 Accuracy 64.09%\n",
            "Epoch 9840/15000 Cost: 0.670539 Accuracy 64.09%\n",
            "Epoch 9850/15000 Cost: 0.670476 Accuracy 64.09%\n",
            "Epoch 9860/15000 Cost: 0.670413 Accuracy 64.09%\n",
            "Epoch 9870/15000 Cost: 0.670350 Accuracy 64.09%\n",
            "Epoch 9880/15000 Cost: 0.670287 Accuracy 64.20%\n",
            "Epoch 9890/15000 Cost: 0.670224 Accuracy 64.20%\n",
            "Epoch 9900/15000 Cost: 0.670161 Accuracy 64.20%\n",
            "Epoch 9910/15000 Cost: 0.670098 Accuracy 64.20%\n",
            "Epoch 9920/15000 Cost: 0.670034 Accuracy 64.20%\n",
            "Epoch 9930/15000 Cost: 0.669971 Accuracy 64.20%\n",
            "Epoch 9940/15000 Cost: 0.669908 Accuracy 64.20%\n",
            "Epoch 9950/15000 Cost: 0.669846 Accuracy 64.20%\n",
            "Epoch 9960/15000 Cost: 0.669783 Accuracy 64.20%\n",
            "Epoch 9970/15000 Cost: 0.669720 Accuracy 64.20%\n",
            "Epoch 9980/15000 Cost: 0.669657 Accuracy 64.20%\n",
            "Epoch 9990/15000 Cost: 0.669594 Accuracy 64.31%\n",
            "Epoch 10000/15000 Cost: 0.669531 Accuracy 64.31%\n",
            "Epoch 10010/15000 Cost: 0.669468 Accuracy 64.31%\n",
            "Epoch 10020/15000 Cost: 0.669405 Accuracy 64.31%\n",
            "Epoch 10030/15000 Cost: 0.669343 Accuracy 64.31%\n",
            "Epoch 10040/15000 Cost: 0.669280 Accuracy 64.31%\n",
            "Epoch 10050/15000 Cost: 0.669217 Accuracy 64.31%\n",
            "Epoch 10060/15000 Cost: 0.669154 Accuracy 64.42%\n",
            "Epoch 10070/15000 Cost: 0.669092 Accuracy 64.42%\n",
            "Epoch 10080/15000 Cost: 0.669029 Accuracy 64.42%\n",
            "Epoch 10090/15000 Cost: 0.668967 Accuracy 64.42%\n",
            "Epoch 10100/15000 Cost: 0.668904 Accuracy 64.42%\n",
            "Epoch 10110/15000 Cost: 0.668841 Accuracy 64.42%\n",
            "Epoch 10120/15000 Cost: 0.668779 Accuracy 64.42%\n",
            "Epoch 10130/15000 Cost: 0.668716 Accuracy 64.42%\n",
            "Epoch 10140/15000 Cost: 0.668654 Accuracy 64.42%\n",
            "Epoch 10150/15000 Cost: 0.668591 Accuracy 64.42%\n",
            "Epoch 10160/15000 Cost: 0.668529 Accuracy 64.42%\n",
            "Epoch 10170/15000 Cost: 0.668466 Accuracy 64.42%\n",
            "Epoch 10180/15000 Cost: 0.668404 Accuracy 64.42%\n",
            "Epoch 10190/15000 Cost: 0.668341 Accuracy 64.42%\n",
            "Epoch 10200/15000 Cost: 0.668279 Accuracy 64.42%\n",
            "Epoch 10210/15000 Cost: 0.668217 Accuracy 64.42%\n",
            "Epoch 10220/15000 Cost: 0.668154 Accuracy 64.42%\n",
            "Epoch 10230/15000 Cost: 0.668092 Accuracy 64.42%\n",
            "Epoch 10240/15000 Cost: 0.668030 Accuracy 64.42%\n",
            "Epoch 10250/15000 Cost: 0.667967 Accuracy 64.42%\n",
            "Epoch 10260/15000 Cost: 0.667905 Accuracy 64.42%\n",
            "Epoch 10270/15000 Cost: 0.667843 Accuracy 64.42%\n",
            "Epoch 10280/15000 Cost: 0.667781 Accuracy 64.42%\n",
            "Epoch 10290/15000 Cost: 0.667719 Accuracy 64.42%\n",
            "Epoch 10300/15000 Cost: 0.667656 Accuracy 64.42%\n",
            "Epoch 10310/15000 Cost: 0.667594 Accuracy 64.42%\n",
            "Epoch 10320/15000 Cost: 0.667532 Accuracy 64.42%\n",
            "Epoch 10330/15000 Cost: 0.667470 Accuracy 64.42%\n",
            "Epoch 10340/15000 Cost: 0.667408 Accuracy 64.42%\n",
            "Epoch 10350/15000 Cost: 0.667346 Accuracy 64.42%\n",
            "Epoch 10360/15000 Cost: 0.667284 Accuracy 64.42%\n",
            "Epoch 10370/15000 Cost: 0.667222 Accuracy 64.53%\n",
            "Epoch 10380/15000 Cost: 0.667160 Accuracy 64.53%\n",
            "Epoch 10390/15000 Cost: 0.667098 Accuracy 64.53%\n",
            "Epoch 10400/15000 Cost: 0.667036 Accuracy 64.53%\n",
            "Epoch 10410/15000 Cost: 0.666974 Accuracy 64.53%\n",
            "Epoch 10420/15000 Cost: 0.666912 Accuracy 64.53%\n",
            "Epoch 10430/15000 Cost: 0.666850 Accuracy 64.53%\n",
            "Epoch 10440/15000 Cost: 0.666789 Accuracy 64.65%\n",
            "Epoch 10450/15000 Cost: 0.666727 Accuracy 64.65%\n",
            "Epoch 10460/15000 Cost: 0.666665 Accuracy 64.65%\n",
            "Epoch 10470/15000 Cost: 0.666603 Accuracy 64.65%\n",
            "Epoch 10480/15000 Cost: 0.666541 Accuracy 64.65%\n",
            "Epoch 10490/15000 Cost: 0.666480 Accuracy 64.65%\n",
            "Epoch 10500/15000 Cost: 0.666418 Accuracy 64.65%\n",
            "Epoch 10510/15000 Cost: 0.666356 Accuracy 64.65%\n",
            "Epoch 10520/15000 Cost: 0.666295 Accuracy 64.65%\n",
            "Epoch 10530/15000 Cost: 0.666233 Accuracy 64.65%\n",
            "Epoch 10540/15000 Cost: 0.666171 Accuracy 64.65%\n",
            "Epoch 10550/15000 Cost: 0.666110 Accuracy 64.65%\n",
            "Epoch 10560/15000 Cost: 0.666048 Accuracy 64.65%\n",
            "Epoch 10570/15000 Cost: 0.665987 Accuracy 64.65%\n",
            "Epoch 10580/15000 Cost: 0.665925 Accuracy 64.65%\n",
            "Epoch 10590/15000 Cost: 0.665864 Accuracy 64.65%\n",
            "Epoch 10600/15000 Cost: 0.665802 Accuracy 64.65%\n",
            "Epoch 10610/15000 Cost: 0.665741 Accuracy 64.65%\n",
            "Epoch 10620/15000 Cost: 0.665679 Accuracy 64.65%\n",
            "Epoch 10630/15000 Cost: 0.665618 Accuracy 64.65%\n",
            "Epoch 10640/15000 Cost: 0.665557 Accuracy 64.65%\n",
            "Epoch 10650/15000 Cost: 0.665495 Accuracy 64.53%\n",
            "Epoch 10660/15000 Cost: 0.665434 Accuracy 64.53%\n",
            "Epoch 10670/15000 Cost: 0.665373 Accuracy 64.53%\n",
            "Epoch 10680/15000 Cost: 0.665311 Accuracy 64.53%\n",
            "Epoch 10690/15000 Cost: 0.665250 Accuracy 64.53%\n",
            "Epoch 10700/15000 Cost: 0.665189 Accuracy 64.53%\n",
            "Epoch 10710/15000 Cost: 0.665128 Accuracy 64.42%\n",
            "Epoch 10720/15000 Cost: 0.665066 Accuracy 64.31%\n",
            "Epoch 10730/15000 Cost: 0.665005 Accuracy 64.31%\n",
            "Epoch 10740/15000 Cost: 0.664944 Accuracy 64.31%\n",
            "Epoch 10750/15000 Cost: 0.664883 Accuracy 64.31%\n",
            "Epoch 10760/15000 Cost: 0.664822 Accuracy 64.31%\n",
            "Epoch 10770/15000 Cost: 0.664761 Accuracy 64.42%\n",
            "Epoch 10780/15000 Cost: 0.664700 Accuracy 64.42%\n",
            "Epoch 10790/15000 Cost: 0.664639 Accuracy 64.42%\n",
            "Epoch 10800/15000 Cost: 0.664578 Accuracy 64.31%\n",
            "Epoch 10810/15000 Cost: 0.664517 Accuracy 64.31%\n",
            "Epoch 10820/15000 Cost: 0.664456 Accuracy 64.31%\n",
            "Epoch 10830/15000 Cost: 0.664395 Accuracy 64.31%\n",
            "Epoch 10840/15000 Cost: 0.664334 Accuracy 64.31%\n",
            "Epoch 10850/15000 Cost: 0.664273 Accuracy 64.31%\n",
            "Epoch 10860/15000 Cost: 0.664212 Accuracy 64.31%\n",
            "Epoch 10870/15000 Cost: 0.664151 Accuracy 64.31%\n",
            "Epoch 10880/15000 Cost: 0.664090 Accuracy 64.31%\n",
            "Epoch 10890/15000 Cost: 0.664029 Accuracy 64.31%\n",
            "Epoch 10900/15000 Cost: 0.663969 Accuracy 64.31%\n",
            "Epoch 10910/15000 Cost: 0.663908 Accuracy 64.31%\n",
            "Epoch 10920/15000 Cost: 0.663847 Accuracy 64.31%\n",
            "Epoch 10930/15000 Cost: 0.663786 Accuracy 64.31%\n",
            "Epoch 10940/15000 Cost: 0.663726 Accuracy 64.31%\n",
            "Epoch 10950/15000 Cost: 0.663665 Accuracy 64.42%\n",
            "Epoch 10960/15000 Cost: 0.663604 Accuracy 64.42%\n",
            "Epoch 10970/15000 Cost: 0.663544 Accuracy 64.42%\n",
            "Epoch 10980/15000 Cost: 0.663483 Accuracy 64.42%\n",
            "Epoch 10990/15000 Cost: 0.663423 Accuracy 64.42%\n",
            "Epoch 11000/15000 Cost: 0.663362 Accuracy 64.42%\n",
            "Epoch 11010/15000 Cost: 0.663302 Accuracy 64.42%\n",
            "Epoch 11020/15000 Cost: 0.663241 Accuracy 64.42%\n",
            "Epoch 11030/15000 Cost: 0.663181 Accuracy 64.42%\n",
            "Epoch 11040/15000 Cost: 0.663120 Accuracy 64.42%\n",
            "Epoch 11050/15000 Cost: 0.663060 Accuracy 64.42%\n",
            "Epoch 11060/15000 Cost: 0.662999 Accuracy 64.42%\n",
            "Epoch 11070/15000 Cost: 0.662939 Accuracy 64.42%\n",
            "Epoch 11080/15000 Cost: 0.662878 Accuracy 64.42%\n",
            "Epoch 11090/15000 Cost: 0.662818 Accuracy 64.31%\n",
            "Epoch 11100/15000 Cost: 0.662758 Accuracy 64.31%\n",
            "Epoch 11110/15000 Cost: 0.662697 Accuracy 64.31%\n",
            "Epoch 11120/15000 Cost: 0.662637 Accuracy 64.31%\n",
            "Epoch 11130/15000 Cost: 0.662577 Accuracy 64.31%\n",
            "Epoch 11140/15000 Cost: 0.662517 Accuracy 64.31%\n",
            "Epoch 11150/15000 Cost: 0.662456 Accuracy 64.31%\n",
            "Epoch 11160/15000 Cost: 0.662396 Accuracy 64.31%\n",
            "Epoch 11170/15000 Cost: 0.662336 Accuracy 64.31%\n",
            "Epoch 11180/15000 Cost: 0.662276 Accuracy 64.31%\n",
            "Epoch 11190/15000 Cost: 0.662216 Accuracy 64.31%\n",
            "Epoch 11200/15000 Cost: 0.662156 Accuracy 64.31%\n",
            "Epoch 11210/15000 Cost: 0.662096 Accuracy 64.31%\n",
            "Epoch 11220/15000 Cost: 0.662036 Accuracy 64.31%\n",
            "Epoch 11230/15000 Cost: 0.661975 Accuracy 64.31%\n",
            "Epoch 11240/15000 Cost: 0.661915 Accuracy 64.31%\n",
            "Epoch 11250/15000 Cost: 0.661855 Accuracy 64.31%\n",
            "Epoch 11260/15000 Cost: 0.661795 Accuracy 64.42%\n",
            "Epoch 11270/15000 Cost: 0.661735 Accuracy 64.42%\n",
            "Epoch 11280/15000 Cost: 0.661676 Accuracy 64.42%\n",
            "Epoch 11290/15000 Cost: 0.661616 Accuracy 64.42%\n",
            "Epoch 11300/15000 Cost: 0.661556 Accuracy 64.31%\n",
            "Epoch 11310/15000 Cost: 0.661496 Accuracy 64.31%\n",
            "Epoch 11320/15000 Cost: 0.661436 Accuracy 64.31%\n",
            "Epoch 11330/15000 Cost: 0.661376 Accuracy 64.20%\n",
            "Epoch 11340/15000 Cost: 0.661317 Accuracy 64.20%\n",
            "Epoch 11350/15000 Cost: 0.661257 Accuracy 64.31%\n",
            "Epoch 11360/15000 Cost: 0.661197 Accuracy 64.31%\n",
            "Epoch 11370/15000 Cost: 0.661137 Accuracy 64.31%\n",
            "Epoch 11380/15000 Cost: 0.661078 Accuracy 64.31%\n",
            "Epoch 11390/15000 Cost: 0.661018 Accuracy 64.31%\n",
            "Epoch 11400/15000 Cost: 0.660958 Accuracy 64.31%\n",
            "Epoch 11410/15000 Cost: 0.660899 Accuracy 64.31%\n",
            "Epoch 11420/15000 Cost: 0.660839 Accuracy 64.31%\n",
            "Epoch 11430/15000 Cost: 0.660779 Accuracy 64.31%\n",
            "Epoch 11440/15000 Cost: 0.660720 Accuracy 64.31%\n",
            "Epoch 11450/15000 Cost: 0.660660 Accuracy 64.31%\n",
            "Epoch 11460/15000 Cost: 0.660601 Accuracy 64.31%\n",
            "Epoch 11470/15000 Cost: 0.660541 Accuracy 64.31%\n",
            "Epoch 11480/15000 Cost: 0.660482 Accuracy 64.31%\n",
            "Epoch 11490/15000 Cost: 0.660422 Accuracy 64.31%\n",
            "Epoch 11500/15000 Cost: 0.660363 Accuracy 64.31%\n",
            "Epoch 11510/15000 Cost: 0.660303 Accuracy 64.31%\n",
            "Epoch 11520/15000 Cost: 0.660244 Accuracy 64.31%\n",
            "Epoch 11530/15000 Cost: 0.660185 Accuracy 64.31%\n",
            "Epoch 11540/15000 Cost: 0.660125 Accuracy 64.31%\n",
            "Epoch 11550/15000 Cost: 0.660066 Accuracy 64.31%\n",
            "Epoch 11560/15000 Cost: 0.660007 Accuracy 64.31%\n",
            "Epoch 11570/15000 Cost: 0.659947 Accuracy 64.42%\n",
            "Epoch 11580/15000 Cost: 0.659888 Accuracy 64.42%\n",
            "Epoch 11590/15000 Cost: 0.659829 Accuracy 64.42%\n",
            "Epoch 11600/15000 Cost: 0.659770 Accuracy 64.42%\n",
            "Epoch 11610/15000 Cost: 0.659711 Accuracy 64.42%\n",
            "Epoch 11620/15000 Cost: 0.659651 Accuracy 64.42%\n",
            "Epoch 11630/15000 Cost: 0.659592 Accuracy 64.42%\n",
            "Epoch 11640/15000 Cost: 0.659533 Accuracy 64.42%\n",
            "Epoch 11650/15000 Cost: 0.659474 Accuracy 64.42%\n",
            "Epoch 11660/15000 Cost: 0.659415 Accuracy 64.42%\n",
            "Epoch 11670/15000 Cost: 0.659356 Accuracy 64.42%\n",
            "Epoch 11680/15000 Cost: 0.659297 Accuracy 64.42%\n",
            "Epoch 11690/15000 Cost: 0.659238 Accuracy 64.42%\n",
            "Epoch 11700/15000 Cost: 0.659179 Accuracy 64.42%\n",
            "Epoch 11710/15000 Cost: 0.659120 Accuracy 64.42%\n",
            "Epoch 11720/15000 Cost: 0.659061 Accuracy 64.42%\n",
            "Epoch 11730/15000 Cost: 0.659002 Accuracy 64.42%\n",
            "Epoch 11740/15000 Cost: 0.658943 Accuracy 64.42%\n",
            "Epoch 11750/15000 Cost: 0.658884 Accuracy 64.42%\n",
            "Epoch 11760/15000 Cost: 0.658826 Accuracy 64.42%\n",
            "Epoch 11770/15000 Cost: 0.658767 Accuracy 64.42%\n",
            "Epoch 11780/15000 Cost: 0.658708 Accuracy 64.42%\n",
            "Epoch 11790/15000 Cost: 0.658649 Accuracy 64.42%\n",
            "Epoch 11800/15000 Cost: 0.658590 Accuracy 64.42%\n",
            "Epoch 11810/15000 Cost: 0.658532 Accuracy 64.42%\n",
            "Epoch 11820/15000 Cost: 0.658473 Accuracy 64.42%\n",
            "Epoch 11830/15000 Cost: 0.658414 Accuracy 64.42%\n",
            "Epoch 11840/15000 Cost: 0.658356 Accuracy 64.53%\n",
            "Epoch 11850/15000 Cost: 0.658297 Accuracy 64.53%\n",
            "Epoch 11860/15000 Cost: 0.658238 Accuracy 64.65%\n",
            "Epoch 11870/15000 Cost: 0.658180 Accuracy 64.65%\n",
            "Epoch 11880/15000 Cost: 0.658121 Accuracy 64.65%\n",
            "Epoch 11890/15000 Cost: 0.658063 Accuracy 64.65%\n",
            "Epoch 11900/15000 Cost: 0.658004 Accuracy 64.76%\n",
            "Epoch 11910/15000 Cost: 0.657946 Accuracy 64.76%\n",
            "Epoch 11920/15000 Cost: 0.657887 Accuracy 64.76%\n",
            "Epoch 11930/15000 Cost: 0.657829 Accuracy 64.76%\n",
            "Epoch 11940/15000 Cost: 0.657770 Accuracy 64.76%\n",
            "Epoch 11950/15000 Cost: 0.657712 Accuracy 64.76%\n",
            "Epoch 11960/15000 Cost: 0.657653 Accuracy 64.76%\n",
            "Epoch 11970/15000 Cost: 0.657595 Accuracy 64.76%\n",
            "Epoch 11980/15000 Cost: 0.657537 Accuracy 64.76%\n",
            "Epoch 11990/15000 Cost: 0.657478 Accuracy 64.76%\n",
            "Epoch 12000/15000 Cost: 0.657420 Accuracy 64.76%\n",
            "Epoch 12010/15000 Cost: 0.657362 Accuracy 64.76%\n",
            "Epoch 12020/15000 Cost: 0.657303 Accuracy 64.76%\n",
            "Epoch 12030/15000 Cost: 0.657245 Accuracy 64.76%\n",
            "Epoch 12040/15000 Cost: 0.657187 Accuracy 64.76%\n",
            "Epoch 12050/15000 Cost: 0.657129 Accuracy 64.76%\n",
            "Epoch 12060/15000 Cost: 0.657071 Accuracy 64.76%\n",
            "Epoch 12070/15000 Cost: 0.657013 Accuracy 64.76%\n",
            "Epoch 12080/15000 Cost: 0.656954 Accuracy 64.65%\n",
            "Epoch 12090/15000 Cost: 0.656896 Accuracy 64.65%\n",
            "Epoch 12100/15000 Cost: 0.656838 Accuracy 64.65%\n",
            "Epoch 12110/15000 Cost: 0.656780 Accuracy 64.65%\n",
            "Epoch 12120/15000 Cost: 0.656722 Accuracy 64.65%\n",
            "Epoch 12130/15000 Cost: 0.656664 Accuracy 64.65%\n",
            "Epoch 12140/15000 Cost: 0.656606 Accuracy 64.65%\n",
            "Epoch 12150/15000 Cost: 0.656548 Accuracy 64.65%\n",
            "Epoch 12160/15000 Cost: 0.656490 Accuracy 64.65%\n",
            "Epoch 12170/15000 Cost: 0.656432 Accuracy 64.65%\n",
            "Epoch 12180/15000 Cost: 0.656374 Accuracy 64.65%\n",
            "Epoch 12190/15000 Cost: 0.656317 Accuracy 64.65%\n",
            "Epoch 12200/15000 Cost: 0.656259 Accuracy 64.65%\n",
            "Epoch 12210/15000 Cost: 0.656201 Accuracy 64.65%\n",
            "Epoch 12220/15000 Cost: 0.656143 Accuracy 64.65%\n",
            "Epoch 12230/15000 Cost: 0.656085 Accuracy 64.65%\n",
            "Epoch 12240/15000 Cost: 0.656027 Accuracy 64.65%\n",
            "Epoch 12250/15000 Cost: 0.655970 Accuracy 64.65%\n",
            "Epoch 12260/15000 Cost: 0.655912 Accuracy 64.65%\n",
            "Epoch 12270/15000 Cost: 0.655854 Accuracy 64.65%\n",
            "Epoch 12280/15000 Cost: 0.655797 Accuracy 64.65%\n",
            "Epoch 12290/15000 Cost: 0.655739 Accuracy 64.65%\n",
            "Epoch 12300/15000 Cost: 0.655681 Accuracy 64.65%\n",
            "Epoch 12310/15000 Cost: 0.655623 Accuracy 64.65%\n",
            "Epoch 12320/15000 Cost: 0.655566 Accuracy 64.65%\n",
            "Epoch 12330/15000 Cost: 0.655508 Accuracy 64.65%\n",
            "Epoch 12340/15000 Cost: 0.655451 Accuracy 64.65%\n",
            "Epoch 12350/15000 Cost: 0.655393 Accuracy 64.65%\n",
            "Epoch 12360/15000 Cost: 0.655336 Accuracy 64.76%\n",
            "Epoch 12370/15000 Cost: 0.655278 Accuracy 64.76%\n",
            "Epoch 12380/15000 Cost: 0.655221 Accuracy 64.76%\n",
            "Epoch 12390/15000 Cost: 0.655163 Accuracy 64.76%\n",
            "Epoch 12400/15000 Cost: 0.655106 Accuracy 64.76%\n",
            "Epoch 12410/15000 Cost: 0.655049 Accuracy 64.76%\n",
            "Epoch 12420/15000 Cost: 0.654991 Accuracy 64.76%\n",
            "Epoch 12430/15000 Cost: 0.654934 Accuracy 64.76%\n",
            "Epoch 12440/15000 Cost: 0.654877 Accuracy 64.76%\n",
            "Epoch 12450/15000 Cost: 0.654819 Accuracy 64.76%\n",
            "Epoch 12460/15000 Cost: 0.654762 Accuracy 64.76%\n",
            "Epoch 12470/15000 Cost: 0.654705 Accuracy 64.76%\n",
            "Epoch 12480/15000 Cost: 0.654647 Accuracy 64.76%\n",
            "Epoch 12490/15000 Cost: 0.654590 Accuracy 64.65%\n",
            "Epoch 12500/15000 Cost: 0.654533 Accuracy 64.65%\n",
            "Epoch 12510/15000 Cost: 0.654476 Accuracy 64.65%\n",
            "Epoch 12520/15000 Cost: 0.654419 Accuracy 64.65%\n",
            "Epoch 12530/15000 Cost: 0.654362 Accuracy 64.65%\n",
            "Epoch 12540/15000 Cost: 0.654305 Accuracy 64.65%\n",
            "Epoch 12550/15000 Cost: 0.654247 Accuracy 64.65%\n",
            "Epoch 12560/15000 Cost: 0.654190 Accuracy 64.76%\n",
            "Epoch 12570/15000 Cost: 0.654133 Accuracy 64.76%\n",
            "Epoch 12580/15000 Cost: 0.654076 Accuracy 64.76%\n",
            "Epoch 12590/15000 Cost: 0.654019 Accuracy 64.76%\n",
            "Epoch 12600/15000 Cost: 0.653962 Accuracy 64.76%\n",
            "Epoch 12610/15000 Cost: 0.653905 Accuracy 64.76%\n",
            "Epoch 12620/15000 Cost: 0.653848 Accuracy 64.76%\n",
            "Epoch 12630/15000 Cost: 0.653791 Accuracy 64.87%\n",
            "Epoch 12640/15000 Cost: 0.653735 Accuracy 64.87%\n",
            "Epoch 12650/15000 Cost: 0.653678 Accuracy 64.87%\n",
            "Epoch 12660/15000 Cost: 0.653621 Accuracy 64.76%\n",
            "Epoch 12670/15000 Cost: 0.653564 Accuracy 64.76%\n",
            "Epoch 12680/15000 Cost: 0.653507 Accuracy 64.09%\n",
            "Epoch 12690/15000 Cost: 0.653451 Accuracy 64.09%\n",
            "Epoch 12700/15000 Cost: 0.653394 Accuracy 64.09%\n",
            "Epoch 12710/15000 Cost: 0.653337 Accuracy 63.97%\n",
            "Epoch 12720/15000 Cost: 0.653280 Accuracy 63.97%\n",
            "Epoch 12730/15000 Cost: 0.653224 Accuracy 63.97%\n",
            "Epoch 12740/15000 Cost: 0.653167 Accuracy 64.09%\n",
            "Epoch 12750/15000 Cost: 0.653110 Accuracy 63.97%\n",
            "Epoch 12760/15000 Cost: 0.653054 Accuracy 63.97%\n",
            "Epoch 12770/15000 Cost: 0.652997 Accuracy 63.97%\n",
            "Epoch 12780/15000 Cost: 0.652941 Accuracy 63.97%\n",
            "Epoch 12790/15000 Cost: 0.652884 Accuracy 63.97%\n",
            "Epoch 12800/15000 Cost: 0.652828 Accuracy 63.64%\n",
            "Epoch 12810/15000 Cost: 0.652771 Accuracy 63.64%\n",
            "Epoch 12820/15000 Cost: 0.652715 Accuracy 63.64%\n",
            "Epoch 12830/15000 Cost: 0.652658 Accuracy 63.64%\n",
            "Epoch 12840/15000 Cost: 0.652602 Accuracy 63.64%\n",
            "Epoch 12850/15000 Cost: 0.652545 Accuracy 63.64%\n",
            "Epoch 12860/15000 Cost: 0.652489 Accuracy 63.64%\n",
            "Epoch 12870/15000 Cost: 0.652432 Accuracy 63.64%\n",
            "Epoch 12880/15000 Cost: 0.652376 Accuracy 63.64%\n",
            "Epoch 12890/15000 Cost: 0.652320 Accuracy 63.64%\n",
            "Epoch 12900/15000 Cost: 0.652263 Accuracy 63.64%\n",
            "Epoch 12910/15000 Cost: 0.652207 Accuracy 63.64%\n",
            "Epoch 12920/15000 Cost: 0.652151 Accuracy 63.64%\n",
            "Epoch 12930/15000 Cost: 0.652095 Accuracy 63.64%\n",
            "Epoch 12940/15000 Cost: 0.652038 Accuracy 63.64%\n",
            "Epoch 12950/15000 Cost: 0.651982 Accuracy 63.52%\n",
            "Epoch 12960/15000 Cost: 0.651926 Accuracy 63.52%\n",
            "Epoch 12970/15000 Cost: 0.651870 Accuracy 63.52%\n",
            "Epoch 12980/15000 Cost: 0.651814 Accuracy 63.52%\n",
            "Epoch 12990/15000 Cost: 0.651758 Accuracy 63.52%\n",
            "Epoch 13000/15000 Cost: 0.651701 Accuracy 63.52%\n",
            "Epoch 13010/15000 Cost: 0.651645 Accuracy 63.52%\n",
            "Epoch 13020/15000 Cost: 0.651589 Accuracy 63.52%\n",
            "Epoch 13030/15000 Cost: 0.651533 Accuracy 63.52%\n",
            "Epoch 13040/15000 Cost: 0.651477 Accuracy 63.52%\n",
            "Epoch 13050/15000 Cost: 0.651421 Accuracy 63.52%\n",
            "Epoch 13060/15000 Cost: 0.651365 Accuracy 63.64%\n",
            "Epoch 13070/15000 Cost: 0.651309 Accuracy 63.64%\n",
            "Epoch 13080/15000 Cost: 0.651254 Accuracy 63.64%\n",
            "Epoch 13090/15000 Cost: 0.651198 Accuracy 63.64%\n",
            "Epoch 13100/15000 Cost: 0.651142 Accuracy 63.75%\n",
            "Epoch 13110/15000 Cost: 0.651086 Accuracy 63.75%\n",
            "Epoch 13120/15000 Cost: 0.651030 Accuracy 63.75%\n",
            "Epoch 13130/15000 Cost: 0.650974 Accuracy 63.75%\n",
            "Epoch 13140/15000 Cost: 0.650919 Accuracy 63.75%\n",
            "Epoch 13150/15000 Cost: 0.650863 Accuracy 63.75%\n",
            "Epoch 13160/15000 Cost: 0.650807 Accuracy 63.75%\n",
            "Epoch 13170/15000 Cost: 0.650751 Accuracy 63.75%\n",
            "Epoch 13180/15000 Cost: 0.650696 Accuracy 63.75%\n",
            "Epoch 13190/15000 Cost: 0.650640 Accuracy 63.75%\n",
            "Epoch 13200/15000 Cost: 0.650584 Accuracy 63.75%\n",
            "Epoch 13210/15000 Cost: 0.650529 Accuracy 63.75%\n",
            "Epoch 13220/15000 Cost: 0.650473 Accuracy 63.75%\n",
            "Epoch 13230/15000 Cost: 0.650417 Accuracy 63.75%\n",
            "Epoch 13240/15000 Cost: 0.650362 Accuracy 63.75%\n",
            "Epoch 13250/15000 Cost: 0.650306 Accuracy 63.75%\n",
            "Epoch 13260/15000 Cost: 0.650251 Accuracy 63.75%\n",
            "Epoch 13270/15000 Cost: 0.650195 Accuracy 63.75%\n",
            "Epoch 13280/15000 Cost: 0.650140 Accuracy 63.75%\n",
            "Epoch 13290/15000 Cost: 0.650084 Accuracy 63.75%\n",
            "Epoch 13300/15000 Cost: 0.650029 Accuracy 63.75%\n",
            "Epoch 13310/15000 Cost: 0.649973 Accuracy 63.75%\n",
            "Epoch 13320/15000 Cost: 0.649918 Accuracy 63.75%\n",
            "Epoch 13330/15000 Cost: 0.649863 Accuracy 63.75%\n",
            "Epoch 13340/15000 Cost: 0.649807 Accuracy 63.75%\n",
            "Epoch 13350/15000 Cost: 0.649752 Accuracy 63.97%\n",
            "Epoch 13360/15000 Cost: 0.649697 Accuracy 63.97%\n",
            "Epoch 13370/15000 Cost: 0.649641 Accuracy 63.97%\n",
            "Epoch 13380/15000 Cost: 0.649586 Accuracy 63.97%\n",
            "Epoch 13390/15000 Cost: 0.649531 Accuracy 63.97%\n",
            "Epoch 13400/15000 Cost: 0.649476 Accuracy 63.97%\n",
            "Epoch 13410/15000 Cost: 0.649420 Accuracy 64.31%\n",
            "Epoch 13420/15000 Cost: 0.649365 Accuracy 65.32%\n",
            "Epoch 13430/15000 Cost: 0.649310 Accuracy 65.32%\n",
            "Epoch 13440/15000 Cost: 0.649255 Accuracy 65.32%\n",
            "Epoch 13450/15000 Cost: 0.649200 Accuracy 65.43%\n",
            "Epoch 13460/15000 Cost: 0.649145 Accuracy 65.43%\n",
            "Epoch 13470/15000 Cost: 0.649090 Accuracy 65.54%\n",
            "Epoch 13480/15000 Cost: 0.649035 Accuracy 65.54%\n",
            "Epoch 13490/15000 Cost: 0.648980 Accuracy 65.54%\n",
            "Epoch 13500/15000 Cost: 0.648925 Accuracy 65.54%\n",
            "Epoch 13510/15000 Cost: 0.648870 Accuracy 65.54%\n",
            "Epoch 13520/15000 Cost: 0.648815 Accuracy 65.54%\n",
            "Epoch 13530/15000 Cost: 0.648760 Accuracy 65.54%\n",
            "Epoch 13540/15000 Cost: 0.648705 Accuracy 65.54%\n",
            "Epoch 13550/15000 Cost: 0.648650 Accuracy 65.66%\n",
            "Epoch 13560/15000 Cost: 0.648595 Accuracy 65.66%\n",
            "Epoch 13570/15000 Cost: 0.648540 Accuracy 65.66%\n",
            "Epoch 13580/15000 Cost: 0.648486 Accuracy 65.66%\n",
            "Epoch 13590/15000 Cost: 0.648431 Accuracy 65.66%\n",
            "Epoch 13600/15000 Cost: 0.648376 Accuracy 65.66%\n",
            "Epoch 13610/15000 Cost: 0.648321 Accuracy 65.66%\n",
            "Epoch 13620/15000 Cost: 0.648267 Accuracy 65.66%\n",
            "Epoch 13630/15000 Cost: 0.648212 Accuracy 65.66%\n",
            "Epoch 13640/15000 Cost: 0.648157 Accuracy 65.66%\n",
            "Epoch 13650/15000 Cost: 0.648102 Accuracy 65.66%\n",
            "Epoch 13660/15000 Cost: 0.648048 Accuracy 65.77%\n",
            "Epoch 13670/15000 Cost: 0.647993 Accuracy 65.77%\n",
            "Epoch 13680/15000 Cost: 0.647939 Accuracy 65.88%\n",
            "Epoch 13690/15000 Cost: 0.647884 Accuracy 65.99%\n",
            "Epoch 13700/15000 Cost: 0.647829 Accuracy 65.99%\n",
            "Epoch 13710/15000 Cost: 0.647775 Accuracy 65.99%\n",
            "Epoch 13720/15000 Cost: 0.647720 Accuracy 66.11%\n",
            "Epoch 13730/15000 Cost: 0.647666 Accuracy 66.11%\n",
            "Epoch 13740/15000 Cost: 0.647611 Accuracy 66.11%\n",
            "Epoch 13750/15000 Cost: 0.647557 Accuracy 66.11%\n",
            "Epoch 13760/15000 Cost: 0.647502 Accuracy 66.11%\n",
            "Epoch 13770/15000 Cost: 0.647448 Accuracy 66.11%\n",
            "Epoch 13780/15000 Cost: 0.647394 Accuracy 65.99%\n",
            "Epoch 13790/15000 Cost: 0.647339 Accuracy 65.99%\n",
            "Epoch 13800/15000 Cost: 0.647285 Accuracy 66.11%\n",
            "Epoch 13810/15000 Cost: 0.647231 Accuracy 66.11%\n",
            "Epoch 13820/15000 Cost: 0.647176 Accuracy 66.11%\n",
            "Epoch 13830/15000 Cost: 0.647122 Accuracy 66.11%\n",
            "Epoch 13840/15000 Cost: 0.647068 Accuracy 66.11%\n",
            "Epoch 13850/15000 Cost: 0.647013 Accuracy 66.11%\n",
            "Epoch 13860/15000 Cost: 0.646959 Accuracy 66.11%\n",
            "Epoch 13870/15000 Cost: 0.646905 Accuracy 66.11%\n",
            "Epoch 13880/15000 Cost: 0.646851 Accuracy 66.11%\n",
            "Epoch 13890/15000 Cost: 0.646797 Accuracy 66.11%\n",
            "Epoch 13900/15000 Cost: 0.646743 Accuracy 66.11%\n",
            "Epoch 13910/15000 Cost: 0.646688 Accuracy 66.22%\n",
            "Epoch 13920/15000 Cost: 0.646634 Accuracy 66.22%\n",
            "Epoch 13930/15000 Cost: 0.646580 Accuracy 66.22%\n",
            "Epoch 13940/15000 Cost: 0.646526 Accuracy 66.22%\n",
            "Epoch 13950/15000 Cost: 0.646472 Accuracy 66.22%\n",
            "Epoch 13960/15000 Cost: 0.646418 Accuracy 66.44%\n",
            "Epoch 13970/15000 Cost: 0.646364 Accuracy 66.44%\n",
            "Epoch 13980/15000 Cost: 0.646310 Accuracy 66.44%\n",
            "Epoch 13990/15000 Cost: 0.646256 Accuracy 66.44%\n",
            "Epoch 14000/15000 Cost: 0.646202 Accuracy 66.44%\n",
            "Epoch 14010/15000 Cost: 0.646148 Accuracy 66.44%\n",
            "Epoch 14020/15000 Cost: 0.646095 Accuracy 66.44%\n",
            "Epoch 14030/15000 Cost: 0.646041 Accuracy 66.44%\n",
            "Epoch 14040/15000 Cost: 0.645987 Accuracy 66.44%\n",
            "Epoch 14050/15000 Cost: 0.645933 Accuracy 66.44%\n",
            "Epoch 14060/15000 Cost: 0.645879 Accuracy 66.55%\n",
            "Epoch 14070/15000 Cost: 0.645825 Accuracy 66.55%\n",
            "Epoch 14080/15000 Cost: 0.645772 Accuracy 66.55%\n",
            "Epoch 14090/15000 Cost: 0.645718 Accuracy 66.67%\n",
            "Epoch 14100/15000 Cost: 0.645664 Accuracy 66.67%\n",
            "Epoch 14110/15000 Cost: 0.645611 Accuracy 66.44%\n",
            "Epoch 14120/15000 Cost: 0.645557 Accuracy 66.44%\n",
            "Epoch 14130/15000 Cost: 0.645503 Accuracy 66.44%\n",
            "Epoch 14140/15000 Cost: 0.645450 Accuracy 66.44%\n",
            "Epoch 14150/15000 Cost: 0.645396 Accuracy 66.44%\n",
            "Epoch 14160/15000 Cost: 0.645343 Accuracy 66.44%\n",
            "Epoch 14170/15000 Cost: 0.645289 Accuracy 66.44%\n",
            "Epoch 14180/15000 Cost: 0.645235 Accuracy 66.44%\n",
            "Epoch 14190/15000 Cost: 0.645182 Accuracy 66.44%\n",
            "Epoch 14200/15000 Cost: 0.645128 Accuracy 66.44%\n",
            "Epoch 14210/15000 Cost: 0.645075 Accuracy 66.44%\n",
            "Epoch 14220/15000 Cost: 0.645022 Accuracy 66.44%\n",
            "Epoch 14230/15000 Cost: 0.644968 Accuracy 66.44%\n",
            "Epoch 14240/15000 Cost: 0.644915 Accuracy 66.44%\n",
            "Epoch 14250/15000 Cost: 0.644861 Accuracy 66.44%\n",
            "Epoch 14260/15000 Cost: 0.644808 Accuracy 66.44%\n",
            "Epoch 14270/15000 Cost: 0.644755 Accuracy 66.55%\n",
            "Epoch 14280/15000 Cost: 0.644701 Accuracy 66.55%\n",
            "Epoch 14290/15000 Cost: 0.644648 Accuracy 66.55%\n",
            "Epoch 14300/15000 Cost: 0.644595 Accuracy 66.55%\n",
            "Epoch 14310/15000 Cost: 0.644542 Accuracy 66.55%\n",
            "Epoch 14320/15000 Cost: 0.644488 Accuracy 66.44%\n",
            "Epoch 14330/15000 Cost: 0.644435 Accuracy 66.44%\n",
            "Epoch 14340/15000 Cost: 0.644382 Accuracy 66.44%\n",
            "Epoch 14350/15000 Cost: 0.644329 Accuracy 66.44%\n",
            "Epoch 14360/15000 Cost: 0.644276 Accuracy 66.44%\n",
            "Epoch 14370/15000 Cost: 0.644222 Accuracy 66.44%\n",
            "Epoch 14380/15000 Cost: 0.644169 Accuracy 66.44%\n",
            "Epoch 14390/15000 Cost: 0.644116 Accuracy 66.44%\n",
            "Epoch 14400/15000 Cost: 0.644063 Accuracy 66.44%\n",
            "Epoch 14410/15000 Cost: 0.644010 Accuracy 66.44%\n",
            "Epoch 14420/15000 Cost: 0.643957 Accuracy 66.44%\n",
            "Epoch 14430/15000 Cost: 0.643904 Accuracy 66.44%\n",
            "Epoch 14440/15000 Cost: 0.643851 Accuracy 66.44%\n",
            "Epoch 14450/15000 Cost: 0.643798 Accuracy 66.44%\n",
            "Epoch 14460/15000 Cost: 0.643745 Accuracy 66.44%\n",
            "Epoch 14470/15000 Cost: 0.643692 Accuracy 66.44%\n",
            "Epoch 14480/15000 Cost: 0.643639 Accuracy 66.44%\n",
            "Epoch 14490/15000 Cost: 0.643586 Accuracy 66.44%\n",
            "Epoch 14500/15000 Cost: 0.643534 Accuracy 66.44%\n",
            "Epoch 14510/15000 Cost: 0.643481 Accuracy 66.44%\n",
            "Epoch 14520/15000 Cost: 0.643428 Accuracy 66.44%\n",
            "Epoch 14530/15000 Cost: 0.643375 Accuracy 66.44%\n",
            "Epoch 14540/15000 Cost: 0.643322 Accuracy 66.44%\n",
            "Epoch 14550/15000 Cost: 0.643270 Accuracy 66.44%\n",
            "Epoch 14560/15000 Cost: 0.643217 Accuracy 66.44%\n",
            "Epoch 14570/15000 Cost: 0.643164 Accuracy 66.44%\n",
            "Epoch 14580/15000 Cost: 0.643111 Accuracy 66.44%\n",
            "Epoch 14590/15000 Cost: 0.643059 Accuracy 66.44%\n",
            "Epoch 14600/15000 Cost: 0.643006 Accuracy 66.44%\n",
            "Epoch 14610/15000 Cost: 0.642954 Accuracy 66.44%\n",
            "Epoch 14620/15000 Cost: 0.642901 Accuracy 66.89%\n",
            "Epoch 14630/15000 Cost: 0.642848 Accuracy 66.89%\n",
            "Epoch 14640/15000 Cost: 0.642796 Accuracy 66.89%\n",
            "Epoch 14650/15000 Cost: 0.642743 Accuracy 66.89%\n",
            "Epoch 14660/15000 Cost: 0.642691 Accuracy 66.89%\n",
            "Epoch 14670/15000 Cost: 0.642638 Accuracy 66.89%\n",
            "Epoch 14680/15000 Cost: 0.642586 Accuracy 66.89%\n",
            "Epoch 14690/15000 Cost: 0.642533 Accuracy 66.78%\n",
            "Epoch 14700/15000 Cost: 0.642481 Accuracy 66.78%\n",
            "Epoch 14710/15000 Cost: 0.642429 Accuracy 66.78%\n",
            "Epoch 14720/15000 Cost: 0.642376 Accuracy 66.78%\n",
            "Epoch 14730/15000 Cost: 0.642324 Accuracy 66.78%\n",
            "Epoch 14740/15000 Cost: 0.642271 Accuracy 66.78%\n",
            "Epoch 14750/15000 Cost: 0.642219 Accuracy 66.89%\n",
            "Epoch 14760/15000 Cost: 0.642167 Accuracy 67.00%\n",
            "Epoch 14770/15000 Cost: 0.642114 Accuracy 67.00%\n",
            "Epoch 14780/15000 Cost: 0.642062 Accuracy 67.00%\n",
            "Epoch 14790/15000 Cost: 0.642010 Accuracy 67.00%\n",
            "Epoch 14800/15000 Cost: 0.641958 Accuracy 67.00%\n",
            "Epoch 14810/15000 Cost: 0.641906 Accuracy 67.00%\n",
            "Epoch 14820/15000 Cost: 0.641853 Accuracy 67.00%\n",
            "Epoch 14830/15000 Cost: 0.641801 Accuracy 67.00%\n",
            "Epoch 14840/15000 Cost: 0.641749 Accuracy 67.34%\n",
            "Epoch 14850/15000 Cost: 0.641697 Accuracy 67.34%\n",
            "Epoch 14860/15000 Cost: 0.641645 Accuracy 67.34%\n",
            "Epoch 14870/15000 Cost: 0.641593 Accuracy 67.34%\n",
            "Epoch 14880/15000 Cost: 0.641541 Accuracy 67.34%\n",
            "Epoch 14890/15000 Cost: 0.641489 Accuracy 67.34%\n",
            "Epoch 14900/15000 Cost: 0.641437 Accuracy 67.34%\n",
            "Epoch 14910/15000 Cost: 0.641385 Accuracy 67.34%\n",
            "Epoch 14920/15000 Cost: 0.641333 Accuracy 67.34%\n",
            "Epoch 14930/15000 Cost: 0.641281 Accuracy 67.34%\n",
            "Epoch 14940/15000 Cost: 0.641229 Accuracy 67.34%\n",
            "Epoch 14950/15000 Cost: 0.641177 Accuracy 67.12%\n",
            "Epoch 14960/15000 Cost: 0.641125 Accuracy 67.12%\n",
            "Epoch 14970/15000 Cost: 0.641073 Accuracy 67.12%\n",
            "Epoch 14980/15000 Cost: 0.641021 Accuracy 67.12%\n",
            "Epoch 14990/15000 Cost: 0.640970 Accuracy 67.12%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_y_pred = model(test_x)"
      ],
      "metadata": {
        "id": "eUl4vz_mPuu0"
      },
      "execution_count": 115,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(test_y_pred)"
      ],
      "metadata": {
        "id": "jUo_AcMOP78N",
        "outputId": "935696eb-9802-45ea-d316-a2719a7bdf05",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 124,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([[1.8010e-05],\n",
            "        [1.4119e-07],\n",
            "        [7.2111e-10],\n",
            "        [2.5338e-04],\n",
            "        [3.7253e-03],\n",
            "        [3.1676e-02],\n",
            "        [1.3320e-04],\n",
            "        [3.4807e-03],\n",
            "        [1.3051e-02],\n",
            "        [8.0938e-03],\n",
            "        [3.0494e-03],\n",
            "        [1.1855e-06],\n",
            "        [8.2784e-01],\n",
            "        [2.0698e-09],\n",
            "        [5.6878e-05],\n",
            "        [9.8982e-03],\n",
            "        [2.0834e-05],\n",
            "        [3.0252e-03],\n",
            "        [2.5129e-04],\n",
            "        [6.2680e-07],\n",
            "        [2.6008e-06],\n",
            "        [1.2430e-01],\n",
            "        [4.6520e-02],\n",
            "        [6.3145e-01],\n",
            "        [1.0000e+00],\n",
            "        [7.8704e-08],\n",
            "        [6.4700e-01],\n",
            "        [1.7422e-03],\n",
            "        [1.2758e-05],\n",
            "        [1.3494e-02],\n",
            "        [2.4995e-07],\n",
            "        [5.2195e-03],\n",
            "        [2.4234e-04],\n",
            "        [3.9505e-02],\n",
            "        [2.1125e-02],\n",
            "        [7.5745e-03],\n",
            "        [2.7585e-02],\n",
            "        [3.3262e-03],\n",
            "        [5.8472e-04],\n",
            "        [4.8972e-01],\n",
            "        [1.1618e-05],\n",
            "        [1.8116e-02],\n",
            "        [1.3185e-06],\n",
            "        [1.6440e-04],\n",
            "        [4.2894e-05],\n",
            "        [4.8538e-04],\n",
            "        [4.1233e-06],\n",
            "        [3.7324e-03],\n",
            "        [5.8220e-06],\n",
            "        [6.1412e-05],\n",
            "        [1.4217e-01],\n",
            "        [6.8091e-04],\n",
            "        [1.6704e-02],\n",
            "        [1.0000e+00],\n",
            "        [9.5130e-03],\n",
            "        [4.5603e-01],\n",
            "        [1.2114e-05],\n",
            "        [4.6985e-04],\n",
            "        [6.0259e-03],\n",
            "        [1.0000e+00],\n",
            "        [9.1619e-03],\n",
            "        [5.7905e-05],\n",
            "        [6.2463e-03],\n",
            "        [2.5677e-03],\n",
            "        [1.0000e+00],\n",
            "        [1.6639e-02],\n",
            "        [1.1330e-02],\n",
            "        [5.7044e-06],\n",
            "        [6.2709e-04],\n",
            "        [9.9998e-01],\n",
            "        [1.2334e-03],\n",
            "        [2.1109e-03],\n",
            "        [1.6005e-04],\n",
            "        [1.4302e-03],\n",
            "        [1.0000e+00],\n",
            "        [1.0000e+00],\n",
            "        [3.1053e-03],\n",
            "        [3.3657e-08],\n",
            "        [1.1410e-04],\n",
            "        [1.2334e-03],\n",
            "        [6.7612e-01],\n",
            "        [8.1489e-01],\n",
            "        [3.9214e-07],\n",
            "        [3.0494e-03],\n",
            "        [4.3132e-03],\n",
            "        [7.6893e-03],\n",
            "        [4.1463e-04],\n",
            "        [9.2966e-03],\n",
            "        [1.7739e-01],\n",
            "        [9.2297e-01],\n",
            "        [3.2104e-03],\n",
            "        [3.0062e-03],\n",
            "        [5.7070e-02],\n",
            "        [3.1053e-03],\n",
            "        [4.2296e-03],\n",
            "        [4.7802e-04],\n",
            "        [1.0438e-08],\n",
            "        [1.1108e-04],\n",
            "        [4.3664e-03],\n",
            "        [2.5794e-05],\n",
            "        [1.9593e-04],\n",
            "        [1.2043e-03],\n",
            "        [3.7324e-03],\n",
            "        [3.2983e-04],\n",
            "        [3.2854e-02],\n",
            "        [9.0253e-04],\n",
            "        [2.6060e-03],\n",
            "        [3.7324e-03],\n",
            "        [3.3576e-03],\n",
            "        [7.8638e-03],\n",
            "        [3.9080e-06],\n",
            "        [3.3103e-02],\n",
            "        [2.0739e-04],\n",
            "        [8.8034e-03],\n",
            "        [9.6519e-01],\n",
            "        [1.5942e-02],\n",
            "        [3.9813e-03],\n",
            "        [9.3565e-01],\n",
            "        [2.4288e-02],\n",
            "        [8.3030e-04],\n",
            "        [1.4803e-01],\n",
            "        [2.8058e-03],\n",
            "        [4.8957e-03],\n",
            "        [1.4829e-04],\n",
            "        [3.7324e-03],\n",
            "        [4.7073e-02],\n",
            "        [1.4437e-03],\n",
            "        [1.0744e-01],\n",
            "        [1.3664e-06],\n",
            "        [7.1212e-04],\n",
            "        [3.7296e-05],\n",
            "        [1.8725e-07],\n",
            "        [1.2066e-01],\n",
            "        [2.9931e-03],\n",
            "        [6.3404e-07],\n",
            "        [6.9583e-04],\n",
            "        [3.9910e-04],\n",
            "        [4.9857e-04],\n",
            "        [1.4825e-03],\n",
            "        [1.1151e-03],\n",
            "        [8.9331e-01],\n",
            "        [9.9828e-01],\n",
            "        [9.9995e-01],\n",
            "        [1.1091e-03],\n",
            "        [5.5301e-06],\n",
            "        [7.4112e-05],\n",
            "        [2.6922e-01],\n",
            "        [1.4877e-03],\n",
            "        [1.8116e-02],\n",
            "        [5.6006e-04],\n",
            "        [9.3948e-01],\n",
            "        [4.7273e-03],\n",
            "        [4.1814e-09],\n",
            "        [3.9567e-05],\n",
            "        [2.9007e-01],\n",
            "        [6.7125e-04],\n",
            "        [1.0000e+00],\n",
            "        [1.4486e-03],\n",
            "        [5.5301e-06],\n",
            "        [1.9045e-03],\n",
            "        [3.2929e-02],\n",
            "        [5.9079e-01],\n",
            "        [7.6195e-04],\n",
            "        [2.7436e-03],\n",
            "        [1.9757e-06],\n",
            "        [2.6992e-03],\n",
            "        [1.5962e-05],\n",
            "        [1.4030e-01],\n",
            "        [4.5268e-02],\n",
            "        [2.3857e-03],\n",
            "        [2.9275e-03],\n",
            "        [3.3192e-04],\n",
            "        [1.5435e-03],\n",
            "        [4.3705e-03],\n",
            "        [1.2683e-04],\n",
            "        [6.3961e-01],\n",
            "        [9.8303e-02],\n",
            "        [2.3546e-06],\n",
            "        [1.0795e-03],\n",
            "        [5.9228e-06],\n",
            "        [1.1410e-04],\n",
            "        [4.4284e-02],\n",
            "        [4.9106e-01],\n",
            "        [3.7324e-03],\n",
            "        [1.0000e+00],\n",
            "        [4.0736e-06],\n",
            "        [1.6175e-02],\n",
            "        [5.2834e-03],\n",
            "        [8.9104e-01],\n",
            "        [2.8567e-06],\n",
            "        [9.1240e-05],\n",
            "        [1.6994e-02],\n",
            "        [1.3937e-01],\n",
            "        [1.4286e-09],\n",
            "        [8.8297e-01],\n",
            "        [2.5203e-05],\n",
            "        [1.0000e+00],\n",
            "        [9.0017e-03],\n",
            "        [1.1207e-03],\n",
            "        [4.5013e-03],\n",
            "        [7.8622e-02],\n",
            "        [9.4848e-01],\n",
            "        [9.9995e-01],\n",
            "        [7.2928e-01],\n",
            "        [5.3633e-04],\n",
            "        [2.5378e-02],\n",
            "        [2.1380e-05],\n",
            "        [7.7532e-04],\n",
            "        [3.9247e-04],\n",
            "        [4.8371e-04],\n",
            "        [2.0662e-04],\n",
            "        [2.7599e-03],\n",
            "        [9.4643e-01],\n",
            "        [9.0155e-09],\n",
            "        [3.5745e-06],\n",
            "        [3.6482e-05],\n",
            "        [3.3484e-02],\n",
            "        [2.2754e-01],\n",
            "        [9.9954e-01],\n",
            "        [3.1053e-03],\n",
            "        [2.1221e-04],\n",
            "        [2.1497e-03],\n",
            "        [2.3320e-03],\n",
            "        [2.0861e-03],\n",
            "        [2.3817e-07],\n",
            "        [3.1218e-02],\n",
            "        [9.9889e-04],\n",
            "        [3.2992e-02],\n",
            "        [3.7374e-06],\n",
            "        [1.2487e-05],\n",
            "        [9.5132e-01],\n",
            "        [1.8054e-02],\n",
            "        [1.3444e-03],\n",
            "        [3.7897e-03],\n",
            "        [3.8571e-03],\n",
            "        [3.0344e-03],\n",
            "        [6.1375e-07],\n",
            "        [4.3683e-03],\n",
            "        [1.4283e-02],\n",
            "        [1.2732e-02],\n",
            "        [1.1768e-07],\n",
            "        [9.6018e-06],\n",
            "        [5.1276e-01],\n",
            "        [3.0464e-03],\n",
            "        [4.4960e-03],\n",
            "        [1.1989e-04],\n",
            "        [8.0284e-03],\n",
            "        [1.4474e-05],\n",
            "        [8.3030e-04],\n",
            "        [1.1042e-02],\n",
            "        [9.8446e-01],\n",
            "        [3.0599e-03],\n",
            "        [9.9867e-01],\n",
            "        [8.2794e-04],\n",
            "        [3.6819e-05],\n",
            "        [2.9275e-03],\n",
            "        [3.7324e-03],\n",
            "        [1.6300e-04],\n",
            "        [9.4082e-03],\n",
            "        [2.0810e-03],\n",
            "        [1.6105e-05],\n",
            "        [2.1006e-03],\n",
            "        [1.5292e-03],\n",
            "        [8.9498e-01],\n",
            "        [1.7176e-04],\n",
            "        [3.0494e-03],\n",
            "        [7.9758e-04],\n",
            "        [2.9275e-03],\n",
            "        [2.7585e-02],\n",
            "        [1.0023e-02],\n",
            "        [6.2275e-04],\n",
            "        [3.7324e-03],\n",
            "        [9.9936e-01],\n",
            "        [6.0226e-02],\n",
            "        [4.3683e-03],\n",
            "        [2.2443e-02],\n",
            "        [1.7748e-04],\n",
            "        [9.9846e-06],\n",
            "        [2.2074e-04],\n",
            "        [1.6196e-03],\n",
            "        [1.5937e-03],\n",
            "        [8.8674e-01],\n",
            "        [3.2992e-02],\n",
            "        [4.9874e-01],\n",
            "        [9.3841e-01],\n",
            "        [7.7621e-06],\n",
            "        [2.8257e-03],\n",
            "        [6.9771e-01],\n",
            "        [4.3705e-03],\n",
            "        [3.1053e-03],\n",
            "        [7.9507e-02],\n",
            "        [1.2292e-04],\n",
            "        [4.3705e-03],\n",
            "        [7.0079e-05],\n",
            "        [1.0129e-05],\n",
            "        [3.3458e-04],\n",
            "        [9.9796e-01],\n",
            "        [1.3494e-02],\n",
            "        [4.3261e-03],\n",
            "        [1.1016e-04],\n",
            "        [3.6102e-05],\n",
            "        [8.9366e-03],\n",
            "        [3.4201e-06],\n",
            "        [7.6559e-04],\n",
            "        [3.2992e-02],\n",
            "        [2.5200e-09],\n",
            "        [9.9945e-01],\n",
            "        [8.5708e-01],\n",
            "        [1.3283e-04],\n",
            "        [6.8429e-07],\n",
            "        [6.9533e-03],\n",
            "        [2.0942e-03],\n",
            "        [2.9362e-03],\n",
            "        [1.0226e-05],\n",
            "        [3.9510e-02],\n",
            "        [1.6026e-02],\n",
            "        [3.5842e-02],\n",
            "        [4.8802e-03],\n",
            "        [2.3028e-04],\n",
            "        [1.0851e-02],\n",
            "        [3.2983e-04],\n",
            "        [6.9405e-04],\n",
            "        [4.9857e-04],\n",
            "        [1.5276e-04],\n",
            "        [9.9999e-01],\n",
            "        [9.1463e-04],\n",
            "        [6.8294e-01],\n",
            "        [9.9425e-04],\n",
            "        [5.7639e-04],\n",
            "        [3.1428e-03],\n",
            "        [7.0576e-06],\n",
            "        [3.7681e-05],\n",
            "        [4.3683e-03],\n",
            "        [2.5733e-02],\n",
            "        [2.3142e-04],\n",
            "        [4.3259e-04],\n",
            "        [5.4580e-05],\n",
            "        [7.8509e-06],\n",
            "        [6.3923e-03],\n",
            "        [6.9405e-04],\n",
            "        [7.0411e-03],\n",
            "        [3.5276e-05],\n",
            "        [9.8630e-01],\n",
            "        [1.0000e+00],\n",
            "        [1.1140e-03],\n",
            "        [1.8369e-02],\n",
            "        [4.9857e-04],\n",
            "        [8.2867e-06],\n",
            "        [1.1052e-03],\n",
            "        [2.9286e-04],\n",
            "        [4.4706e-04],\n",
            "        [5.3633e-04],\n",
            "        [9.2435e-01],\n",
            "        [7.1899e-05],\n",
            "        [9.7772e-01],\n",
            "        [2.7120e-07],\n",
            "        [1.6247e-07],\n",
            "        [3.0434e-03],\n",
            "        [3.7324e-03],\n",
            "        [2.0485e-04],\n",
            "        [8.7220e-01],\n",
            "        [4.0412e-02],\n",
            "        [2.9286e-04],\n",
            "        [2.5338e-04],\n",
            "        [1.3011e-01],\n",
            "        [9.9045e-01],\n",
            "        [7.6930e-03],\n",
            "        [8.3054e-02],\n",
            "        [2.7997e-04],\n",
            "        [2.8354e-04],\n",
            "        [1.9789e-03],\n",
            "        [9.9603e-01],\n",
            "        [1.8106e-08],\n",
            "        [6.5354e-07],\n",
            "        [6.9833e-05],\n",
            "        [1.0000e+00],\n",
            "        [1.3002e-03],\n",
            "        [2.6332e-03],\n",
            "        [7.3356e-07],\n",
            "        [8.8645e-01],\n",
            "        [3.7324e-03],\n",
            "        [4.1608e-04],\n",
            "        [9.5337e-03],\n",
            "        [1.2473e-02],\n",
            "        [4.4717e-03],\n",
            "        [5.1099e-01],\n",
            "        [6.8935e-04],\n",
            "        [5.4118e-09],\n",
            "        [2.5843e-03],\n",
            "        [6.0175e-01],\n",
            "        [9.4376e-01],\n",
            "        [1.8526e-06],\n",
            "        [2.5594e-01],\n",
            "        [1.6084e-07],\n",
            "        [3.5034e-04],\n",
            "        [6.8576e-01],\n",
            "        [8.0721e-04],\n",
            "        [7.2319e-04],\n",
            "        [1.4401e-03],\n",
            "        [6.4729e-05],\n",
            "        [9.9988e-01],\n",
            "        [1.1555e-05],\n",
            "        [5.7464e-01],\n",
            "        [3.8801e-01],\n",
            "        [5.1230e-06],\n",
            "        [7.7784e-03],\n",
            "        [8.4189e-04],\n",
            "        [9.9933e-01],\n",
            "        [3.2882e-02],\n",
            "        [8.3110e-01],\n",
            "        [3.2992e-02],\n",
            "        [7.8901e-02],\n",
            "        [2.2734e-04],\n",
            "        [3.1053e-03],\n",
            "        [3.8866e-01],\n",
            "        [3.0875e-06],\n",
            "        [3.1053e-03],\n",
            "        [4.3680e-03]], grad_fn=<SigmoidBackward0>)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sub['Survived'] = torch.Tensor.detach(test_y_pred)\n",
        "sub.to_csv('베이스 라인 .csv', index = False)"
      ],
      "metadata": {
        "id": "74ftZpgoPf5T"
      },
      "execution_count": 127,
      "outputs": []
    }
  ]
}