3_Hidden_2_Conv_ConvolutionalNetworksModifiedDropout_NoBoost_20Epochs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "3 Hidden 2 Conv ConvolutionalNetworksModifiedDropout_NoBoost_20Epochs.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python2",
      "display_name": "Python 2"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/preritt/CS231n-/blob/master/3_Hidden_2_Conv_ConvolutionalNetworksModifiedDropout_NoBoost_20Epochs.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "sYiV4RkdLECs",
        "colab_type": "code",
        "outputId": "3be87f05-d8da-4e95-da1e-903b25774c03",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "%cd .."
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "xoaDYXkvLITX",
        "colab_type": "code",
        "outputId": "9017d49b-58cf-4afb-b678-542985ced934",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "ls"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34massignment2\u001b[0m/  assignment2ModifiedDropout.zip  \u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "14rvTP-1f197",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!rm assignment2ModifiedDropout.zip"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "zyDcCBTeK-fk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!rm -rf assignment2"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QQagLPf1f_uS",
        "colab_type": "code",
        "outputId": "a0f8a642-b93d-40a3-c597-9511f5b3863e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "pwd"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "u'/content'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        }
      ]
    },
    {
      "metadata": {
        "id": "epO1jU1WgEDp",
        "colab_type": "code",
        "outputId": "8e2ba73c-b119-4cdb-a194-7c307b30e8db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "ls"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34massignment2\u001b[0m/  assignment2.zip  layers.py  \u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "WvpGkXRnFDHc",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import zipfile\n",
        "with zipfile.ZipFile(\"assignment2ModifiedDropout_NoBoost.zip\",\"r\") as zip_ref:\n",
        "    zip_ref.extractall()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_DUyji4bGIEW",
        "colab_type": "code",
        "outputId": "3b3a7d39-0f95-4169-a33c-313b560350c4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "%cd assignment2"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/assignment2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "kzD3TNRRHybd",
        "colab_type": "code",
        "outputId": "af03e1af-16cc-4466-9dd9-9c90dd86173b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "%cd cs231n"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/assignment2/cs231n\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "oyEIyCZVGMri",
        "colab_type": "code",
        "outputId": "2add344a-4e8d-4e7b-c41f-9f74dc4270b7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "pwd"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "u'/content/assignment2/cs231n'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "metadata": {
        "id": "O9cf0V5PGOPQ",
        "colab_type": "code",
        "outputId": "e2102cb3-dff8-4b1d-a395-bac0e4346888",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "!python setup.py build_ext --inplace"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "running build_ext\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "3qPo7KaUH7tS",
        "colab_type": "code",
        "outputId": "2d807113-7e77-4985-b4cb-2e833795f513",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "%cd .."
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/assignment2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Ag0449_siA8T",
        "colab_type": "code",
        "outputId": "b96704df-6cd4-4c5d-bd8e-0dfbd5f85e28",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "pwd"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "u'/content/assignment2'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "metadata": {
        "id": "bTha130wkwdr",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from cs231n.classifiers.cnn import *"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "3W_zUjy4ICTM",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# As usual, a bit of setup\n",
        "\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from cs231n.classifiers.cnn import *\n",
        "from cs231n.data_utils import get_CIFAR10_data\n",
        "from cs231n.gradient_check import eval_numerical_gradient_array, eval_numerical_gradient\n",
        "from cs231n.layers import *\n",
        "from cs231n.fast_layers import *\n",
        "from cs231n.solver import Solver\n",
        "\n",
        "%matplotlib inline\n",
        "plt.rcParams['figure.figsize'] = (10.0, 8.0) # set default size of plots\n",
        "plt.rcParams['image.interpolation'] = 'nearest'\n",
        "plt.rcParams['image.cmap'] = 'gray'\n",
        "\n",
        "# for auto-reloading external modules\n",
        "# see http://stackoverflow.com/questions/1907993/autoreload-of-modules-in-ipython\n",
        "%load_ext autoreload\n",
        "%autoreload 2\n",
        "\n",
        "def rel_error(x, y):\n",
        "  \"\"\" returns relative error \"\"\"\n",
        "  return np.max(np.abs(x - y) / (np.maximum(1e-8, np.abs(x) + np.abs(y))))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "cvPZDQC8Jagz",
        "colab_type": "code",
        "outputId": "2e4b09a9-8c44-4936-b44c-7bf2eaa6994e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "cell_type": "code",
      "source": [
        "# Load the (preprocessed) CIFAR10 data.\n",
        "\n",
        "data = get_CIFAR10_data()\n",
        "for k, v in data.iteritems():\n",
        "  print '%s: ' % k, v.shape"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "X_val:  (1000, 3, 32, 32)\n",
            "X_train:  (49000, 3, 32, 32)\n",
            "X_test:  (1000, 3, 32, 32)\n",
            "y_val:  (1000,)\n",
            "y_train:  (49000,)\n",
            "y_test:  (1000,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "S9MX7dQqM8yD",
        "colab_type": "code",
        "outputId": "034568c9-49df-4853-afc5-fd36ad4ecfde",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "x_shape = (2, 3, 4, 4)\n",
        "w_shape = (3, 3, 4, 4)\n",
        "x = np.linspace(-0.1, 0.5, num=np.prod(x_shape)).reshape(x_shape)\n",
        "w = np.linspace(-0.2, 0.3, num=np.prod(w_shape)).reshape(w_shape)\n",
        "b = np.linspace(-0.1, 0.2, num=3)\n",
        "\n",
        "conv_param = {'stride': 2, 'pad': 1}\n",
        "out, _ = conv_forward_naive(x, w, b, conv_param)\n",
        "correct_out = np.array([[[[[-0.08759809, -0.10987781],\n",
        "                           [-0.18387192, -0.2109216 ]],\n",
        "                          [[ 0.21027089,  0.21661097],\n",
        "                           [ 0.22847626,  0.23004637]],\n",
        "                          [[ 0.50813986,  0.54309974],\n",
        "                           [ 0.64082444,  0.67101435]]],\n",
        "                         [[[-0.98053589, -1.03143541],\n",
        "                           [-1.19128892, -1.24695841]],\n",
        "                          [[ 0.69108355,  0.66880383],\n",
        "                           [ 0.59480972,  0.56776003]],\n",
        "                          [[ 2.36270298,  2.36904306],\n",
        "                           [ 2.38090835,  2.38247847]]]]])\n",
        "\n",
        "# Compare your output to ours; difference should be around 1e-8\n",
        "print 'Testing conv_forward_naive'\n",
        "print 'difference: ', rel_error(out, correct_out)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing conv_forward_naive\n",
            "difference:  2.2121476575931688e-08\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "3yzl2o1aJn_x",
        "colab_type": "code",
        "outputId": "632074f9-763a-4e43-929b-378136844ff4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 668
        }
      },
      "cell_type": "code",
      "source": [
        "from scipy.misc import imread, imresize\n",
        "\n",
        "kitten, puppy = imread('kitten.jpg'), imread('puppy.jpg')\n",
        "# kitten is wide, and puppy is already square\n",
        "d = kitten.shape[1] - kitten.shape[0]\n",
        "kitten_cropped = kitten[:, d/2:-d/2, :]\n",
        "\n",
        "img_size = 200   # Make this smaller if it runs too slow\n",
        "x = np.zeros((2, 3, img_size, img_size))\n",
        "x[0, :, :, :] = imresize(puppy, (img_size, img_size)).transpose((2, 0, 1))\n",
        "x[1, :, :, :] = imresize(kitten_cropped, (img_size, img_size)).transpose((2, 0, 1))\n",
        "\n",
        "# Set up a convolutional weights holding 2 filters, each 3x3\n",
        "w = np.zeros((2, 3, 3, 3))\n",
        "\n",
        "# The first filter converts the image to grayscale.\n",
        "# Set up the red, green, and blue channels of the filter.\n",
        "w[0, 0, :, :] = [[0, 0, 0], [0, 0.3, 0], [0, 0, 0]]\n",
        "w[0, 1, :, :] = [[0, 0, 0], [0, 0.6, 0], [0, 0, 0]]\n",
        "w[0, 2, :, :] = [[0, 0, 0], [0, 0.1, 0], [0, 0, 0]]\n",
        "\n",
        "# Second filter detects horizontal edges in the blue channel.\n",
        "w[1, 2, :, :] = [[1, 2, 1], [0, 0, 0], [-1, -2, -1]]\n",
        "\n",
        "# Vector of biases. We don't need any bias for the grayscale\n",
        "# filter, but for the edge detection filter we want to add 128\n",
        "# to each output so that nothing is negative.\n",
        "b = np.array([0, 128])\n",
        "\n",
        "# Compute the result of convolving each input in x with each filter in w,\n",
        "# offsetting by b, and storing the results in out.\n",
        "out, _ = conv_forward_naive(x, w, b, {'stride': 1, 'pad': 1})\n",
        "\n",
        "def imshow_noax(img, normalize=True):\n",
        "    \"\"\" Tiny helper to show images as uint8 and remove axis labels \"\"\"\n",
        "    if normalize:\n",
        "        img_max, img_min = np.max(img), np.min(img)\n",
        "        img = 255.0 * (img - img_min) / (img_max - img_min)\n",
        "    plt.imshow(img.astype('uint8'))\n",
        "    plt.gca().axis('off')\n",
        "\n",
        "# Show the original images and the results of the conv operation\n",
        "plt.subplot(2, 3, 1)\n",
        "imshow_noax(puppy, normalize=False)\n",
        "plt.title('Original image')\n",
        "plt.subplot(2, 3, 2)\n",
        "imshow_noax(out[0, 0])\n",
        "plt.title('Grayscale')\n",
        "plt.subplot(2, 3, 3)\n",
        "imshow_noax(out[0, 1])\n",
        "plt.title('Edges')\n",
        "plt.subplot(2, 3, 4)\n",
        "imshow_noax(kitten_cropped, normalize=False)\n",
        "plt.subplot(2, 3, 5)\n",
        "imshow_noax(out[1, 0])\n",
        "plt.subplot(2, 3, 6)\n",
        "imshow_noax(out[1, 1])\n",
        "plt.show()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python2.7/dist-packages/ipykernel_launcher.py:3: DeprecationWarning: `imread` is deprecated!\n",
            "`imread` is deprecated in SciPy 1.0.0, and will be removed in 1.2.0.\n",
            "Use ``imageio.imread`` instead.\n",
            "  This is separate from the ipykernel package so we can avoid doing imports until\n",
            "/usr/local/lib/python2.7/dist-packages/ipykernel_launcher.py:10: DeprecationWarning: `imresize` is deprecated!\n",
            "`imresize` is deprecated in SciPy 1.0.0, and will be removed in 1.2.0.\n",
            "Use ``skimage.transform.resize`` instead.\n",
            "  # Remove the CWD from sys.path while we load stuff.\n",
            "/usr/local/lib/python2.7/dist-packages/ipykernel_launcher.py:11: DeprecationWarning: `imresize` is deprecated!\n",
            "`imresize` is deprecated in SciPy 1.0.0, and will be removed in 1.2.0.\n",
            "Use ``skimage.transform.resize`` instead.\n",
            "  # This is added back by InteractiveShellApp.init_path()\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl0AAAG/CAYAAABi5mI/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
            "text/plain": [
              "<Figure size 720x576 with 6 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "u4XonbrpNNTr",
        "colab_type": "code",
        "outputId": "21aca810-03f3-4aa7-bb81-9f66b1c002f3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "cell_type": "code",
      "source": [
        "x = np.random.randn(4, 3, 5, 5)\n",
        "w = np.random.randn(2, 3, 3, 3)\n",
        "b = np.random.randn(2,)\n",
        "dout = np.random.randn(4, 2, 5, 5)\n",
        "conv_param = {'stride': 1, 'pad': 1}\n",
        " \n",
        "dx_num = eval_numerical_gradient_array(lambda x: conv_forward_naive(x, w, b, conv_param)[0], x, dout)\n",
        "dw_num = eval_numerical_gradient_array(lambda w: conv_forward_naive(x, w, b, conv_param)[0], w, dout)\n",
        "db_num = eval_numerical_gradient_array(lambda b: conv_forward_naive(x, w, b, conv_param)[0], b, dout)\n",
        "\n",
        "out, cache = conv_forward_naive(x, w, b, conv_param)\n",
        "dx, dw, db = conv_backward_naive(dout, cache)\n",
        "\n",
        "# Your errors should be around 1e-9'\n",
        "\n",
        "print 'Testing conv_backward_naive function'\n",
        "print 'dx error: ', rel_error(dx, dx_num)\n",
        "print 'dw error: ', rel_error(dw, dw_num)\n",
        "print 'db error: ', rel_error(db, db_num)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing conv_backward_naive function\n",
            "dx error:  3.4469012869252418e-09\n",
            "dw error:  4.4822253552497033e-10\n",
            "db error:  1.3125168129366248e-11\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "2u0Kvz0bNRaz",
        "colab_type": "code",
        "outputId": "bbb65a58-8a3b-4ab7-ff56-64ea06f50270",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "x_shape = (2, 3, 4, 4)\n",
        "x = np.linspace(-0.3, 0.4, num=np.prod(x_shape)).reshape(x_shape)\n",
        "pool_param = {'pool_width': 2, 'pool_height': 2, 'stride': 2}\n",
        "\n",
        "out, _ = max_pool_forward_naive(x, pool_param)\n",
        "\n",
        "correct_out = np.array([[[[-0.26315789, -0.24842105],\n",
        "                          [-0.20421053, -0.18947368]],\n",
        "                         [[-0.14526316, -0.13052632],\n",
        "                          [-0.08631579, -0.07157895]],\n",
        "                         [[-0.02736842, -0.01263158],\n",
        "                          [ 0.03157895,  0.04631579]]],\n",
        "                        [[[ 0.09052632,  0.10526316],\n",
        "                          [ 0.14947368,  0.16421053]],\n",
        "                         [[ 0.20842105,  0.22315789],\n",
        "                          [ 0.26736842,  0.28210526]],\n",
        "                         [[ 0.32631579,  0.34105263],\n",
        "                          [ 0.38526316,  0.4       ]]]])\n",
        "\n",
        "#print out\n",
        "# Compare your output with ours. Difference should be around 1e-8.\n",
        "print 'Testing max_pool_forward_naive function:'\n",
        "print 'difference: ', rel_error(out, correct_out)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing max_pool_forward_naive function:\n",
            "difference:  4.1666665157267834e-08\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "RW2s5v5UNVpm",
        "colab_type": "code",
        "outputId": "99777977-d3d3-4662-c45d-aa5acceced75",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "x = np.random.randn(3, 2, 8, 8)\n",
        "dout = np.random.randn(3, 2, 4, 4)\n",
        "pool_param = {'pool_height': 2, 'pool_width': 2, 'stride': 2}\n",
        "\n",
        "dx_num = eval_numerical_gradient_array(lambda x: max_pool_forward_naive(x, pool_param)[0], x, dout)\n",
        "\n",
        "out, cache = max_pool_forward_naive(x, pool_param)\n",
        "dx = max_pool_backward_naive(dout, cache)\n",
        "\n",
        "# Your error should be around 1e-12\n",
        "print 'Testing max_pool_backward_naive function:'\n",
        "print 'dx error: ', rel_error(dx, dx_num)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing max_pool_backward_naive function:\n",
            "dx error:  3.275610742465456e-12\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "JmsMTNE6NYVK",
        "colab_type": "code",
        "outputId": "6120f65b-2371-4485-b7c9-0d18e7b23f95",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        }
      },
      "cell_type": "code",
      "source": [
        "from cs231n.fast_layers import conv_forward_fast, conv_backward_fast\n",
        "from time import time\n",
        "\n",
        "x = np.random.randn(100, 3, 31, 31)\n",
        "w = np.random.randn(25, 3, 3, 3)\n",
        "b = np.random.randn(25,)\n",
        "dout = np.random.randn(100, 25, 16, 16)\n",
        "conv_param = {'stride': 2, 'pad': 1}\n",
        "\n",
        "t0 = time()\n",
        "out_naive, cache_naive = conv_forward_naive(x, w, b, conv_param)\n",
        "t1 = time()\n",
        "out_fast, cache_fast = conv_forward_fast(x, w, b, conv_param)\n",
        "t2 = time()\n",
        "\n",
        "print 'Testing conv_forward_fast:'\n",
        "print 'Naive: %fs' % (t1 - t0)\n",
        "print 'Fast: %fs' % (t2 - t1)\n",
        "print 'Speedup: %fx' % ((t1 - t0) / (t2 - t1))\n",
        "print 'Difference: ', rel_error(out_naive, out_fast)\n",
        "\n",
        "t0 = time()\n",
        "dx_naive, dw_naive, db_naive = conv_backward_naive(dout, cache_naive)\n",
        "t1 = time()\n",
        "dx_fast, dw_fast, db_fast = conv_backward_fast(dout, cache_fast)\n",
        "t2 = time()\n",
        "\n",
        "print '\\nTesting conv_backward_fast:'\n",
        "print 'Naive: %fs' % (t1 - t0)\n",
        "print 'Fast: %fs' % (t2 - t1)\n",
        "print 'Speedup: %fx' % ((t1 - t0) / (t2 - t1))\n",
        "print 'dx difference: ', rel_error(dx_naive, dx_fast)\n",
        "print 'dw difference: ', rel_error(dw_naive, dw_fast)\n",
        "print 'db difference: ', rel_error(db_naive, db_fast)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing conv_forward_fast:\n",
            "Naive: 0.094182s\n",
            "Fast: 0.015364s\n",
            "Speedup: 6.130072x\n",
            "Difference:  3.232951145417351e-12\n",
            "\n",
            "Testing conv_backward_fast:\n",
            "Naive: 15.997482s\n",
            "Fast: 0.013803s\n",
            "Speedup: 1158.985439x\n",
            "dx difference:  5.398911720382192e-12\n",
            "dw difference:  1.1181716975620626e-14\n",
            "db difference:  0.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "oBdL-hnRNmOW",
        "colab_type": "code",
        "outputId": "e61dd49f-3dae-451d-c600-1c5583143277",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "cell_type": "code",
      "source": [
        "from cs231n.fast_layers import max_pool_forward_fast, max_pool_backward_fast\n",
        "\n",
        "x = np.random.randn(100, 3, 32, 32)\n",
        "dout = np.random.randn(100, 3, 16, 16)\n",
        "pool_param = {'pool_height': 2, 'pool_width': 2, 'stride': 2}\n",
        "\n",
        "t0 = time()\n",
        "out_naive, cache_naive = max_pool_forward_naive(x, pool_param)\n",
        "t1 = time()\n",
        "out_fast, cache_fast = max_pool_forward_fast(x, pool_param)\n",
        "t2 = time()\n",
        "\n",
        "print 'Testing pool_forward_fast:'\n",
        "print 'Naive: %fs' % (t1 - t0)\n",
        "print 'fast: %fs' % (t2 - t1)\n",
        "print 'speedup: %fx' % ((t1 - t0) / (t2 - t1))\n",
        "print 'difference: ', rel_error(out_naive, out_fast)\n",
        "\n",
        "t0 = time()\n",
        "dx_naive = max_pool_backward_naive(dout, cache_naive)\n",
        "t1 = time()\n",
        "dx_fast = max_pool_backward_fast(dout, cache_fast)\n",
        "t2 = time()\n",
        "\n",
        "print '\\nTesting pool_backward_fast:'\n",
        "print 'Naive: %fs' % (t1 - t0)\n",
        "print 'speedup: %fx' % ((t1 - t0) / (t2 - t1))\n",
        "print 'dx difference: ', rel_error(dx_naive, dx_fast)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing pool_forward_fast:\n",
            "Naive: 0.301458s\n",
            "fast: 0.002809s\n",
            "speedup: 107.325949x\n",
            "difference:  0.0\n",
            "\n",
            "Testing pool_backward_fast:\n",
            "Naive: 0.650990s\n",
            "speedup: 42.724029x\n",
            "dx difference:  0.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "PVGe0v5oIHRy",
        "colab_type": "code",
        "outputId": "a1531248-6a92-4a98-a54b-280344f0ce93",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "cell_type": "code",
      "source": [
        "from cs231n.layer_utils import conv_relu_pool_forward, conv_relu_pool_backward\n",
        "\n",
        "x = np.random.randn(2, 3, 16, 16)\n",
        "w = np.random.randn(3, 3, 3, 3)\n",
        "b = np.random.randn(3,)\n",
        "dout = np.random.randn(2, 3, 8, 8)\n",
        "conv_param = {'stride': 1, 'pad': 1}\n",
        "pool_param = {'pool_height': 2, 'pool_width': 2, 'stride': 2}\n",
        "\n",
        "out, cache = conv_relu_pool_forward(x, w, b, conv_param, pool_param)\n",
        "dx, dw, db = conv_relu_pool_backward(dout, cache)\n",
        "\n",
        "dx_num = eval_numerical_gradient_array(lambda x: conv_relu_pool_forward(x, w, b, conv_param, pool_param)[0], x, dout)\n",
        "dw_num = eval_numerical_gradient_array(lambda w: conv_relu_pool_forward(x, w, b, conv_param, pool_param)[0], w, dout)\n",
        "db_num = eval_numerical_gradient_array(lambda b: conv_relu_pool_forward(x, w, b, conv_param, pool_param)[0], b, dout)\n",
        "\n",
        "print 'Testing conv_relu_pool'\n",
        "print 'dx error: ', rel_error(dx_num, dx)\n",
        "print 'dw error: ', rel_error(dw_num, dw)\n",
        "print 'db error: ', rel_error(db_num, db)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing conv_relu_pool\n",
            "dx error:  7.67613189087017e-08\n",
            "dw error:  8.863792304852461e-10\n",
            "db error:  5.982406804560325e-10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "yZmAdnlQN1jb",
        "colab_type": "code",
        "outputId": "b4da5093-2d30-402d-a3bd-46a85ee908c3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "cell_type": "code",
      "source": [
        "from cs231n.layer_utils import conv_relu_forward, conv_relu_backward\n",
        "\n",
        "x = np.random.randn(2, 3, 8, 8)\n",
        "w = np.random.randn(3, 3, 3, 3)\n",
        "b = np.random.randn(3,)\n",
        "dout = np.random.randn(2, 3, 8, 8)\n",
        "conv_param = {'stride': 1, 'pad': 1}\n",
        "\n",
        "out, cache = conv_relu_forward(x, w, b, conv_param)\n",
        "dx, dw, db = conv_relu_backward(dout, cache)\n",
        "\n",
        "dx_num = eval_numerical_gradient_array(lambda x: conv_relu_forward(x, w, b, conv_param)[0], x, dout)\n",
        "dw_num = eval_numerical_gradient_array(lambda w: conv_relu_forward(x, w, b, conv_param)[0], w, dout)\n",
        "db_num = eval_numerical_gradient_array(lambda b: conv_relu_forward(x, w, b, conv_param)[0], b, dout)\n",
        "\n",
        "print 'Testing conv_relu:'\n",
        "print 'dx error: ', rel_error(dx_num, dx)\n",
        "print 'dw error: ', rel_error(dw_num, dw)\n",
        "print 'db error: ', rel_error(db_num, db)"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Testing conv_relu:\n",
            "dx error:  9.489289007652193e-09\n",
            "dw error:  3.452879138504745e-10\n",
            "db error:  9.262895624495962e-12\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "H9f-0jXBN6yq",
        "colab_type": "code",
        "outputId": "0d9ad0d4-2bcd-4b24-c0af-ff8b28a78831",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "model = ThreeLayerConvNet()\n",
        "\n",
        "N = 50\n",
        "X = np.random.randn(N, 3, 32, 32)\n",
        "y = np.random.randint(10, size=N)\n",
        "\n",
        "loss, grads = model.loss(X, y)\n",
        "print 'Initial loss (no regularization): ', loss\n",
        "\n",
        "model.reg = 0.5\n",
        "loss, grads = model.loss(X, y)\n",
        "print 'Initial loss (with regularization): ', loss"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Initial loss (no regularization):  2.3025831552453178\n",
            "Initial loss (with regularization):  2.5091171236939567\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "kwrS2BiHOBTP",
        "colab_type": "code",
        "outputId": "e19b791d-5adf-498e-8382-75ce4624caed",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "cell_type": "code",
      "source": [
        "num_inputs = 2\n",
        "input_dim = (3, 16, 16)\n",
        "reg = 0.0\n",
        "num_classes = 10\n",
        "X = np.random.randn(num_inputs, *input_dim)\n",
        "y = np.random.randint(num_classes, size=num_inputs)\n",
        "\n",
        "model = ThreeLayerConvNet(num_filters=3, filter_size=3,\n",
        "                          input_dim=input_dim, hidden_dim=7,\n",
        "                          dtype=np.float64)\n",
        "loss, grads = model.loss(X, y)\n",
        "for param_name in sorted(grads):\n",
        "    f = lambda _: model.loss(X, y)[0]\n",
        "    param_grad_num = eval_numerical_gradient(f, model.params[param_name], verbose=False, h=1e-6)\n",
        "    e = rel_error(param_grad_num, grads[param_name])\n",
        "    print '%s max relative error: %e' % (param_name, rel_error(param_grad_num, grads[param_name]))"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "W1 max relative error: 4.451713e-04\n",
            "W2 max relative error: 1.178048e-02\n",
            "W3 max relative error: 5.062994e-05\n",
            "b1 max relative error: 1.175036e-05\n",
            "b2 max relative error: 1.680469e-07\n",
            "b3 max relative error: 8.221541e-10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "MV_Ob77kOMnE",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "num_train = 100\n",
        "small_data = {\n",
        "  'X_train': data['X_train'][:num_train],\n",
        "  'y_train': data['y_train'][:num_train],\n",
        "  'X_val': data['X_val'],\n",
        "  'y_val': data['y_val'],\n",
        "}\n",
        "\n",
        "model = ThreeLayerConvNet(weight_scale=1e-2)\n",
        "\n",
        "solver = Solver(model, small_data,\n",
        "                num_epochs=20, batch_size=50,\n",
        "                update_rule='adam',\n",
        "                optim_config={\n",
        "                  'learning_rate': 1e-3,\n",
        "                },\n",
        "                verbose=True, print_every=1)\n",
        "solver.train()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "uHC3cjHDOlHV",
        "colab_type": "code",
        "outputId": "bc1b4bfb-40da-4ded-c5a8-9a3b6e0a2fae",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        }
      },
      "cell_type": "code",
      "source": [
        "plt.subplot(2, 1, 1)\n",
        "plt.plot(solver.loss_history, 'o')\n",
        "plt.xlabel('iteration')\n",
        "plt.ylabel('loss')\n",
        "\n",
        "plt.subplot(2, 1, 2)\n",
        "plt.plot(solver.train_acc_history, '-o')\n",
        "plt.plot(solver.val_acc_history, '-o')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.xlabel('epoch')\n",
        "plt.ylabel('accuracy')\n",
        "plt.show()"
      ],
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAHgCAYAAACbywggAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 720x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "5h9n6o66OoTN",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model = ThreeLayerConvNet(weight_scale=0.001, hidden_dim=500, reg=0.001)\n",
        "\n",
        "solver = Solver(model, data,\n",
        "                num_epochs=1, batch_size=50,\n",
        "                update_rule='adam',\n",
        "                optim_config={\n",
        "                  'learning_rate': 1e-3,\n",
        "                },\n",
        "                verbose=True, print_every=20)\n",
        "solver.train()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "1DyA-_YlOqQD",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "_1LfjHWhPJ6v",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Ymndp0SzPJ1K",
        "colab_type": "code",
        "outputId": "5ea42ba3-b097-4193-9c0d-307126a6ea10",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 554
        }
      },
      "cell_type": "code",
      "source": [
        "model = MultiLayerConvNet(weight_scale=0.001, hidden_dim=[800,400,200], reg=0.001, dropout = 0.5)\n",
        "\n",
        "solver = Solver(model, data,\n",
        "                num_epochs=5, batch_size=50,\n",
        "                update_rule='adam',\n",
        "                optim_config={\n",
        "                  'learning_rate': 1e-3,\n",
        "                },\n",
        "                verbose=True, print_every=20)\n",
        "solver.train()"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "error",
          "ename": "IndexError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m\u001b[0m",
            "\u001b[0;31mIndexError\u001b[0mTraceback (most recent call last)",
            "\u001b[0;32m<ipython-input-40-ceac0e443b0b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      8\u001b[0m                 },\n\u001b[1;32m      9\u001b[0m                 verbose=True, print_every=20)\n\u001b[0;32m---> 10\u001b[0;31m \u001b[0msolver\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/content/assignment2/cs231n/solver.py\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    225\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    226\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mt\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mxrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnum_iterations\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 227\u001b[0;31m       \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    228\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    229\u001b[0m       \u001b[0;31m# Maybe print training loss\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/content/assignment2/cs231n/solver.py\u001b[0m in \u001b[0;36m_step\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    163\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    164\u001b[0m     \u001b[0;31m# Compute loss and gradient\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 165\u001b[0;31m     \u001b[0mloss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgrads\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    166\u001b[0m     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloss_history\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    167\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/content/assignment2/cs231n/classifiers/cnn.py\u001b[0m in \u001b[0;36mloss\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m    347\u001b[0m                                 W3, b3,gamma3, beta3,self.bn_params[2], self.dropout_param)\n\u001b[1;32m    348\u001b[0m     hidden_2, hidden_2_cache = affine_batchnorm_relu_dropout_forward(hidden_1, \n\u001b[0;32m--> 349\u001b[0;31m                             W4, b4,gamma4, beta4,self.bn_params[4], self.dropout_param)\n\u001b[0m\u001b[1;32m    350\u001b[0m     hidden_3, hidden_3_cache = affine_batchnorm_relu_dropout_forward(hidden_2, \n\u001b[1;32m    351\u001b[0m                             W5, b5,gamma5, beta5,self.bn_params[5], self.dropout_param)\n",
            "\u001b[0;31mIndexError\u001b[0m: list index out of range"
          ]
        }
      ]
    },
    {
      "metadata": {
        "id": "IIqA8sJflioF",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from cs231n.classifiers.cnn import *"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "OlH5VfGrHHfn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17034
        },
        "outputId": "206b327f-172c-4ffb-e3bb-65017d17c0f9"
      },
      "cell_type": "code",
      "source": [
        "model = MultiLayerConvNet(weight_scale=0.001, hidden_dim=[800,400,200], reg=0.001, dropout = 0.5)\n",
        "\n",
        "solver = Solver(model, data,\n",
        "                num_epochs=20, batch_size=50,\n",
        "                update_rule='adam',\n",
        "                optim_config={\n",
        "                  'learning_rate': 1e-3,\n",
        "                },\n",
        "                verbose=True, print_every=20)\n",
        "solver.train()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(Iteration 1 / 19600) loss: 2.306457\n",
            "(Epoch 0 / 20) train acc: 0.090000; val_acc: 0.102000\n",
            "(Iteration 21 / 19600) loss: 2.138022\n",
            "(Iteration 41 / 19600) loss: 2.034290\n",
            "(Iteration 61 / 19600) loss: 1.803936\n",
            "(Iteration 81 / 19600) loss: 1.671160\n",
            "(Iteration 101 / 19600) loss: 1.840541\n",
            "(Iteration 121 / 19600) loss: 1.848102\n",
            "(Iteration 141 / 19600) loss: 1.745273\n",
            "(Iteration 161 / 19600) loss: 1.599023\n",
            "(Iteration 181 / 19600) loss: 1.716362\n",
            "(Iteration 201 / 19600) loss: 1.643015\n",
            "(Iteration 221 / 19600) loss: 1.592673\n",
            "(Iteration 241 / 19600) loss: 1.716145\n",
            "(Iteration 261 / 19600) loss: 1.849197\n",
            "(Iteration 281 / 19600) loss: 1.644026\n",
            "(Iteration 301 / 19600) loss: 1.552789\n",
            "(Iteration 321 / 19600) loss: 1.592270\n",
            "(Iteration 341 / 19600) loss: 1.635873\n",
            "(Iteration 361 / 19600) loss: 1.645391\n",
            "(Iteration 381 / 19600) loss: 1.478154\n",
            "(Iteration 401 / 19600) loss: 1.508860\n",
            "(Iteration 421 / 19600) loss: 1.496415\n",
            "(Iteration 441 / 19600) loss: 1.643705\n",
            "(Iteration 461 / 19600) loss: 1.393044\n",
            "(Iteration 481 / 19600) loss: 1.353871\n",
            "(Iteration 501 / 19600) loss: 1.343089\n",
            "(Iteration 521 / 19600) loss: 1.268810\n",
            "(Iteration 541 / 19600) loss: 1.502538\n",
            "(Iteration 561 / 19600) loss: 1.342124\n",
            "(Iteration 581 / 19600) loss: 1.653630\n",
            "(Iteration 601 / 19600) loss: 1.590276\n",
            "(Iteration 621 / 19600) loss: 1.355718\n",
            "(Iteration 641 / 19600) loss: 1.659397\n",
            "(Iteration 661 / 19600) loss: 1.327723\n",
            "(Iteration 681 / 19600) loss: 1.202785\n",
            "(Iteration 701 / 19600) loss: 1.282211\n",
            "(Iteration 721 / 19600) loss: 1.386838\n",
            "(Iteration 741 / 19600) loss: 1.321971\n",
            "(Iteration 761 / 19600) loss: 1.423083\n",
            "(Iteration 781 / 19600) loss: 1.420593\n",
            "(Iteration 801 / 19600) loss: 1.710751\n",
            "(Iteration 821 / 19600) loss: 1.317807\n",
            "(Iteration 841 / 19600) loss: 1.797174\n",
            "(Iteration 861 / 19600) loss: 1.678480\n",
            "(Iteration 881 / 19600) loss: 1.690071\n",
            "(Iteration 901 / 19600) loss: 1.468382\n",
            "(Iteration 921 / 19600) loss: 1.322057\n",
            "(Iteration 941 / 19600) loss: 1.417411\n",
            "(Iteration 961 / 19600) loss: 1.641409\n",
            "(Epoch 1 / 20) train acc: 0.598000; val_acc: 0.583000\n",
            "(Iteration 981 / 19600) loss: 1.176523\n",
            "(Iteration 1001 / 19600) loss: 1.401114\n",
            "(Iteration 1021 / 19600) loss: 1.561240\n",
            "(Iteration 1041 / 19600) loss: 1.272028\n",
            "(Iteration 1061 / 19600) loss: 1.201045\n",
            "(Iteration 1081 / 19600) loss: 1.495322\n",
            "(Iteration 1101 / 19600) loss: 1.255897\n",
            "(Iteration 1121 / 19600) loss: 1.362238\n",
            "(Iteration 1141 / 19600) loss: 1.080240\n",
            "(Iteration 1161 / 19600) loss: 1.677371\n",
            "(Iteration 1181 / 19600) loss: 1.431579\n",
            "(Iteration 1201 / 19600) loss: 1.527663\n",
            "(Iteration 1221 / 19600) loss: 1.296827\n",
            "(Iteration 1241 / 19600) loss: 1.793211\n",
            "(Iteration 1261 / 19600) loss: 1.387353\n",
            "(Iteration 1281 / 19600) loss: 1.263502\n",
            "(Iteration 1301 / 19600) loss: 1.174392\n",
            "(Iteration 1321 / 19600) loss: 1.309374\n",
            "(Iteration 1341 / 19600) loss: 1.192847\n",
            "(Iteration 1361 / 19600) loss: 1.417071\n",
            "(Iteration 1381 / 19600) loss: 1.515000\n",
            "(Iteration 1401 / 19600) loss: 1.383086\n",
            "(Iteration 1421 / 19600) loss: 1.484903\n",
            "(Iteration 1441 / 19600) loss: 1.281546\n",
            "(Iteration 1461 / 19600) loss: 1.485056\n",
            "(Iteration 1481 / 19600) loss: 1.515379\n",
            "(Iteration 1501 / 19600) loss: 1.401650\n",
            "(Iteration 1521 / 19600) loss: 1.222639\n",
            "(Iteration 1541 / 19600) loss: 1.209057\n",
            "(Iteration 1561 / 19600) loss: 1.261876\n",
            "(Iteration 1581 / 19600) loss: 1.221047\n",
            "(Iteration 1601 / 19600) loss: 1.666722\n",
            "(Iteration 1621 / 19600) loss: 1.364867\n",
            "(Iteration 1641 / 19600) loss: 1.518259\n",
            "(Iteration 1661 / 19600) loss: 1.411834\n",
            "(Iteration 1681 / 19600) loss: 1.542738\n",
            "(Iteration 1701 / 19600) loss: 1.441150\n",
            "(Iteration 1721 / 19600) loss: 1.350522\n",
            "(Iteration 1741 / 19600) loss: 1.551773\n",
            "(Iteration 1761 / 19600) loss: 1.285368\n",
            "(Iteration 1781 / 19600) loss: 1.335490\n",
            "(Iteration 1801 / 19600) loss: 1.298719\n",
            "(Iteration 1821 / 19600) loss: 1.364006\n",
            "(Iteration 1841 / 19600) loss: 1.328002\n",
            "(Iteration 1861 / 19600) loss: 1.124754\n",
            "(Iteration 1881 / 19600) loss: 1.258281\n",
            "(Iteration 1901 / 19600) loss: 1.214962\n",
            "(Iteration 1921 / 19600) loss: 1.234186\n",
            "(Iteration 1941 / 19600) loss: 1.519598\n",
            "(Epoch 2 / 20) train acc: 0.680000; val_acc: 0.669000\n",
            "(Iteration 1961 / 19600) loss: 1.255217\n",
            "(Iteration 1981 / 19600) loss: 1.398447\n",
            "(Iteration 2001 / 19600) loss: 1.431021\n",
            "(Iteration 2021 / 19600) loss: 1.507295\n",
            "(Iteration 2041 / 19600) loss: 1.253991\n",
            "(Iteration 2061 / 19600) loss: 1.449845\n",
            "(Iteration 2081 / 19600) loss: 0.866254\n",
            "(Iteration 2101 / 19600) loss: 1.252149\n",
            "(Iteration 2121 / 19600) loss: 0.955661\n",
            "(Iteration 2141 / 19600) loss: 1.318196\n",
            "(Iteration 2161 / 19600) loss: 1.292421\n",
            "(Iteration 2181 / 19600) loss: 1.161854\n",
            "(Iteration 2201 / 19600) loss: 1.309496\n",
            "(Iteration 2221 / 19600) loss: 1.422671\n",
            "(Iteration 2241 / 19600) loss: 1.365181\n",
            "(Iteration 2261 / 19600) loss: 1.120369\n",
            "(Iteration 2281 / 19600) loss: 1.251626\n",
            "(Iteration 2301 / 19600) loss: 1.199785\n",
            "(Iteration 2321 / 19600) loss: 1.295448\n",
            "(Iteration 2341 / 19600) loss: 1.300343\n",
            "(Iteration 2361 / 19600) loss: 1.406144\n",
            "(Iteration 2381 / 19600) loss: 1.471136\n",
            "(Iteration 2401 / 19600) loss: 1.560533\n",
            "(Iteration 2421 / 19600) loss: 1.291359\n",
            "(Iteration 2441 / 19600) loss: 1.511176\n",
            "(Iteration 2461 / 19600) loss: 1.353980\n",
            "(Iteration 2481 / 19600) loss: 1.257750\n",
            "(Iteration 2501 / 19600) loss: 1.215811\n",
            "(Iteration 2521 / 19600) loss: 1.095445\n",
            "(Iteration 2541 / 19600) loss: 1.301773\n",
            "(Iteration 2561 / 19600) loss: 1.418992\n",
            "(Iteration 2581 / 19600) loss: 1.151318\n",
            "(Iteration 2601 / 19600) loss: 1.443315\n",
            "(Iteration 2621 / 19600) loss: 1.542860\n",
            "(Iteration 2641 / 19600) loss: 1.361856\n",
            "(Iteration 2661 / 19600) loss: 1.252225\n",
            "(Iteration 2681 / 19600) loss: 1.166889\n",
            "(Iteration 2701 / 19600) loss: 1.287939\n",
            "(Iteration 2721 / 19600) loss: 1.195493\n",
            "(Iteration 2741 / 19600) loss: 1.144798\n",
            "(Iteration 2761 / 19600) loss: 1.245561\n",
            "(Iteration 2781 / 19600) loss: 1.424134\n",
            "(Iteration 2801 / 19600) loss: 1.185856\n",
            "(Iteration 2821 / 19600) loss: 1.520355\n",
            "(Iteration 2841 / 19600) loss: 1.247596\n",
            "(Iteration 2861 / 19600) loss: 1.335150\n",
            "(Iteration 2881 / 19600) loss: 1.363179\n",
            "(Iteration 2901 / 19600) loss: 1.285146\n",
            "(Iteration 2921 / 19600) loss: 1.102543\n",
            "(Epoch 3 / 20) train acc: 0.750000; val_acc: 0.685000\n",
            "(Iteration 2941 / 19600) loss: 1.218490\n",
            "(Iteration 2961 / 19600) loss: 1.160127\n",
            "(Iteration 2981 / 19600) loss: 1.462801\n",
            "(Iteration 3001 / 19600) loss: 1.101656\n",
            "(Iteration 3021 / 19600) loss: 1.309820\n",
            "(Iteration 3041 / 19600) loss: 1.453138\n",
            "(Iteration 3061 / 19600) loss: 1.260975\n",
            "(Iteration 3081 / 19600) loss: 1.203264\n",
            "(Iteration 3101 / 19600) loss: 1.311191\n",
            "(Iteration 3121 / 19600) loss: 1.339458\n",
            "(Iteration 3141 / 19600) loss: 1.396989\n",
            "(Iteration 3161 / 19600) loss: 1.239972\n",
            "(Iteration 3181 / 19600) loss: 1.433962\n",
            "(Iteration 3201 / 19600) loss: 1.517912\n",
            "(Iteration 3221 / 19600) loss: 1.069177\n",
            "(Iteration 3241 / 19600) loss: 1.266479\n",
            "(Iteration 3261 / 19600) loss: 1.441062\n",
            "(Iteration 3281 / 19600) loss: 1.207377\n",
            "(Iteration 3301 / 19600) loss: 1.042838\n",
            "(Iteration 3321 / 19600) loss: 1.343062\n",
            "(Iteration 3341 / 19600) loss: 1.325573\n",
            "(Iteration 3361 / 19600) loss: 1.153413\n",
            "(Iteration 3381 / 19600) loss: 1.297265\n",
            "(Iteration 3401 / 19600) loss: 1.186889\n",
            "(Iteration 3421 / 19600) loss: 1.226624\n",
            "(Iteration 3441 / 19600) loss: 1.244071\n",
            "(Iteration 3461 / 19600) loss: 0.986159\n",
            "(Iteration 3481 / 19600) loss: 1.615581\n",
            "(Iteration 3501 / 19600) loss: 1.362155\n",
            "(Iteration 3521 / 19600) loss: 1.104792\n",
            "(Iteration 3541 / 19600) loss: 1.105355\n",
            "(Iteration 3561 / 19600) loss: 1.439604\n",
            "(Iteration 3581 / 19600) loss: 1.377992\n",
            "(Iteration 3601 / 19600) loss: 1.597235\n",
            "(Iteration 3621 / 19600) loss: 1.313040\n",
            "(Iteration 3641 / 19600) loss: 1.363018\n",
            "(Iteration 3661 / 19600) loss: 1.751143\n",
            "(Iteration 3681 / 19600) loss: 1.184312\n",
            "(Iteration 3701 / 19600) loss: 1.446045\n",
            "(Iteration 3721 / 19600) loss: 1.430385\n",
            "(Iteration 3741 / 19600) loss: 1.150113\n",
            "(Iteration 3761 / 19600) loss: 1.208592\n",
            "(Iteration 3781 / 19600) loss: 1.105354\n",
            "(Iteration 3801 / 19600) loss: 1.253730\n",
            "(Iteration 3821 / 19600) loss: 1.187993\n",
            "(Iteration 3841 / 19600) loss: 1.341972\n",
            "(Iteration 3861 / 19600) loss: 1.225337\n",
            "(Iteration 3881 / 19600) loss: 1.201998\n",
            "(Iteration 3901 / 19600) loss: 1.043936\n",
            "(Epoch 4 / 20) train acc: 0.710000; val_acc: 0.684000\n",
            "(Iteration 3921 / 19600) loss: 1.048929\n",
            "(Iteration 3941 / 19600) loss: 1.094947\n",
            "(Iteration 3961 / 19600) loss: 1.218290\n",
            "(Iteration 3981 / 19600) loss: 1.192482\n",
            "(Iteration 4001 / 19600) loss: 1.189760\n",
            "(Iteration 4021 / 19600) loss: 1.153619\n",
            "(Iteration 4041 / 19600) loss: 1.233468\n",
            "(Iteration 4061 / 19600) loss: 1.215413\n",
            "(Iteration 4081 / 19600) loss: 1.287190\n",
            "(Iteration 4101 / 19600) loss: 1.221753\n",
            "(Iteration 4121 / 19600) loss: 1.076660\n",
            "(Iteration 4141 / 19600) loss: 1.757223\n",
            "(Iteration 4161 / 19600) loss: 1.149315\n",
            "(Iteration 4181 / 19600) loss: 1.221967\n",
            "(Iteration 4201 / 19600) loss: 1.170729\n",
            "(Iteration 4221 / 19600) loss: 1.103331\n",
            "(Iteration 4241 / 19600) loss: 1.294915\n",
            "(Iteration 4261 / 19600) loss: 1.115869\n",
            "(Iteration 4281 / 19600) loss: 1.266702\n",
            "(Iteration 4301 / 19600) loss: 1.253824\n",
            "(Iteration 4321 / 19600) loss: 1.219919\n",
            "(Iteration 4341 / 19600) loss: 1.062629\n",
            "(Iteration 4361 / 19600) loss: 1.217111\n",
            "(Iteration 4381 / 19600) loss: 1.085985\n",
            "(Iteration 4401 / 19600) loss: 0.945266\n",
            "(Iteration 4421 / 19600) loss: 1.121074\n",
            "(Iteration 4441 / 19600) loss: 1.161388\n",
            "(Iteration 4461 / 19600) loss: 0.984713\n",
            "(Iteration 4481 / 19600) loss: 1.371250\n",
            "(Iteration 4501 / 19600) loss: 1.317384\n",
            "(Iteration 4521 / 19600) loss: 1.312467\n",
            "(Iteration 4541 / 19600) loss: 1.491291\n",
            "(Iteration 4561 / 19600) loss: 1.087674\n",
            "(Iteration 4581 / 19600) loss: 1.190199\n",
            "(Iteration 4601 / 19600) loss: 1.440629\n",
            "(Iteration 4621 / 19600) loss: 1.381991\n",
            "(Iteration 4641 / 19600) loss: 1.302379\n",
            "(Iteration 4661 / 19600) loss: 1.577307\n",
            "(Iteration 4681 / 19600) loss: 1.367185\n",
            "(Iteration 4701 / 19600) loss: 1.142781\n",
            "(Iteration 4721 / 19600) loss: 1.089019\n",
            "(Iteration 4741 / 19600) loss: 1.030227\n",
            "(Iteration 4761 / 19600) loss: 1.307866\n",
            "(Iteration 4781 / 19600) loss: 1.472090\n",
            "(Iteration 4801 / 19600) loss: 1.102385\n",
            "(Iteration 4821 / 19600) loss: 1.001114\n",
            "(Iteration 4841 / 19600) loss: 1.326103\n",
            "(Iteration 4861 / 19600) loss: 1.131677\n",
            "(Iteration 4881 / 19600) loss: 1.003950\n",
            "(Epoch 5 / 20) train acc: 0.763000; val_acc: 0.706000\n",
            "(Iteration 4901 / 19600) loss: 1.299571\n",
            "(Iteration 4921 / 19600) loss: 1.178309\n",
            "(Iteration 4941 / 19600) loss: 1.069040\n",
            "(Iteration 4961 / 19600) loss: 1.504603\n",
            "(Iteration 4981 / 19600) loss: 1.069824\n",
            "(Iteration 5001 / 19600) loss: 1.249028\n",
            "(Iteration 5021 / 19600) loss: 1.273430\n",
            "(Iteration 5041 / 19600) loss: 1.236161\n",
            "(Iteration 5061 / 19600) loss: 1.192180\n",
            "(Iteration 5081 / 19600) loss: 1.246180\n",
            "(Iteration 5101 / 19600) loss: 1.357926\n",
            "(Iteration 5121 / 19600) loss: 1.166757\n",
            "(Iteration 5141 / 19600) loss: 1.351847\n",
            "(Iteration 5161 / 19600) loss: 1.242377\n",
            "(Iteration 5181 / 19600) loss: 1.524939\n",
            "(Iteration 5201 / 19600) loss: 1.045297\n",
            "(Iteration 5221 / 19600) loss: 1.252799\n",
            "(Iteration 5241 / 19600) loss: 1.176092\n",
            "(Iteration 5261 / 19600) loss: 1.446899\n",
            "(Iteration 5281 / 19600) loss: 1.326647\n",
            "(Iteration 5301 / 19600) loss: 1.322163\n",
            "(Iteration 5321 / 19600) loss: 1.101660\n",
            "(Iteration 5341 / 19600) loss: 1.024085\n",
            "(Iteration 5361 / 19600) loss: 1.399057\n",
            "(Iteration 5381 / 19600) loss: 1.179474\n",
            "(Iteration 5401 / 19600) loss: 1.024984\n",
            "(Iteration 5421 / 19600) loss: 1.266203\n",
            "(Iteration 5441 / 19600) loss: 1.337934\n",
            "(Iteration 5461 / 19600) loss: 1.044998\n",
            "(Iteration 5481 / 19600) loss: 1.063635\n",
            "(Iteration 5501 / 19600) loss: 1.283092\n",
            "(Iteration 5521 / 19600) loss: 1.402561\n",
            "(Iteration 5541 / 19600) loss: 1.267005\n",
            "(Iteration 5561 / 19600) loss: 1.139976\n",
            "(Iteration 5581 / 19600) loss: 1.066447\n",
            "(Iteration 5601 / 19600) loss: 1.167255\n",
            "(Iteration 5621 / 19600) loss: 0.970459\n",
            "(Iteration 5641 / 19600) loss: 1.193544\n",
            "(Iteration 5661 / 19600) loss: 1.237762\n",
            "(Iteration 5681 / 19600) loss: 0.954120\n",
            "(Iteration 5701 / 19600) loss: 1.036637\n",
            "(Iteration 5721 / 19600) loss: 1.082275\n",
            "(Iteration 5741 / 19600) loss: 1.413260\n",
            "(Iteration 5761 / 19600) loss: 1.163980\n",
            "(Iteration 5781 / 19600) loss: 1.160035\n",
            "(Iteration 5801 / 19600) loss: 1.279580\n",
            "(Iteration 5821 / 19600) loss: 1.281063\n",
            "(Iteration 5841 / 19600) loss: 1.285536\n",
            "(Iteration 5861 / 19600) loss: 1.267437\n",
            "(Epoch 6 / 20) train acc: 0.748000; val_acc: 0.733000\n",
            "(Iteration 5881 / 19600) loss: 1.020978\n",
            "(Iteration 5901 / 19600) loss: 1.536156\n",
            "(Iteration 5921 / 19600) loss: 1.115890\n",
            "(Iteration 5941 / 19600) loss: 1.171962\n",
            "(Iteration 5961 / 19600) loss: 1.198075\n",
            "(Iteration 5981 / 19600) loss: 1.165811\n",
            "(Iteration 6001 / 19600) loss: 1.272139\n",
            "(Iteration 6021 / 19600) loss: 1.491865\n",
            "(Iteration 6041 / 19600) loss: 1.111425\n",
            "(Iteration 6061 / 19600) loss: 1.470485\n",
            "(Iteration 6081 / 19600) loss: 1.175479\n",
            "(Iteration 6101 / 19600) loss: 1.215031\n",
            "(Iteration 6121 / 19600) loss: 1.141616\n",
            "(Iteration 6141 / 19600) loss: 1.185086\n",
            "(Iteration 6161 / 19600) loss: 1.011725\n",
            "(Iteration 6181 / 19600) loss: 1.136785\n",
            "(Iteration 6201 / 19600) loss: 1.377503\n",
            "(Iteration 6221 / 19600) loss: 1.223463\n",
            "(Iteration 6241 / 19600) loss: 1.029282\n",
            "(Iteration 6261 / 19600) loss: 1.230175\n",
            "(Iteration 6281 / 19600) loss: 1.072371\n",
            "(Iteration 6301 / 19600) loss: 1.306855\n",
            "(Iteration 6321 / 19600) loss: 0.976336\n",
            "(Iteration 6341 / 19600) loss: 1.255650\n",
            "(Iteration 6361 / 19600) loss: 1.471131\n",
            "(Iteration 6381 / 19600) loss: 1.457443\n",
            "(Iteration 6401 / 19600) loss: 0.963666\n",
            "(Iteration 6421 / 19600) loss: 1.271178\n",
            "(Iteration 6441 / 19600) loss: 1.359486\n",
            "(Iteration 6461 / 19600) loss: 1.327720\n",
            "(Iteration 6481 / 19600) loss: 1.331168\n",
            "(Iteration 6501 / 19600) loss: 1.015393\n",
            "(Iteration 6521 / 19600) loss: 1.442450\n",
            "(Iteration 6541 / 19600) loss: 0.960016\n",
            "(Iteration 6561 / 19600) loss: 1.326302\n",
            "(Iteration 6581 / 19600) loss: 1.104816\n",
            "(Iteration 6601 / 19600) loss: 1.143284\n",
            "(Iteration 6621 / 19600) loss: 1.035354\n",
            "(Iteration 6641 / 19600) loss: 1.028363\n",
            "(Iteration 6661 / 19600) loss: 1.267521\n",
            "(Iteration 6681 / 19600) loss: 1.131730\n",
            "(Iteration 6701 / 19600) loss: 1.027575\n",
            "(Iteration 6721 / 19600) loss: 1.285183\n",
            "(Iteration 6741 / 19600) loss: 1.204557\n",
            "(Iteration 6761 / 19600) loss: 1.259483\n",
            "(Iteration 6781 / 19600) loss: 1.473003\n",
            "(Iteration 6801 / 19600) loss: 1.143852\n",
            "(Iteration 6821 / 19600) loss: 1.012569\n",
            "(Iteration 6841 / 19600) loss: 1.094423\n",
            "(Epoch 7 / 20) train acc: 0.774000; val_acc: 0.718000\n",
            "(Iteration 6861 / 19600) loss: 1.090981\n",
            "(Iteration 6881 / 19600) loss: 1.527590\n",
            "(Iteration 6901 / 19600) loss: 1.104855\n",
            "(Iteration 6921 / 19600) loss: 1.090416\n",
            "(Iteration 6941 / 19600) loss: 1.324616\n",
            "(Iteration 6961 / 19600) loss: 1.052219\n",
            "(Iteration 6981 / 19600) loss: 1.163800\n",
            "(Iteration 7001 / 19600) loss: 1.030178\n",
            "(Iteration 7021 / 19600) loss: 1.035226\n",
            "(Iteration 7041 / 19600) loss: 1.181348\n",
            "(Iteration 7061 / 19600) loss: 1.032837\n",
            "(Iteration 7081 / 19600) loss: 1.331225\n",
            "(Iteration 7101 / 19600) loss: 1.274433\n",
            "(Iteration 7121 / 19600) loss: 1.206116\n",
            "(Iteration 7141 / 19600) loss: 1.228866\n",
            "(Iteration 7161 / 19600) loss: 1.285961\n",
            "(Iteration 7181 / 19600) loss: 1.347963\n",
            "(Iteration 7201 / 19600) loss: 1.224885\n",
            "(Iteration 7221 / 19600) loss: 1.100658\n",
            "(Iteration 7241 / 19600) loss: 1.161875\n",
            "(Iteration 7261 / 19600) loss: 1.074639\n",
            "(Iteration 7281 / 19600) loss: 1.348457\n",
            "(Iteration 7301 / 19600) loss: 1.109110\n",
            "(Iteration 7321 / 19600) loss: 1.156075\n",
            "(Iteration 7341 / 19600) loss: 1.218428\n",
            "(Iteration 7361 / 19600) loss: 1.025705\n",
            "(Iteration 7381 / 19600) loss: 1.253545\n",
            "(Iteration 7401 / 19600) loss: 0.965777\n",
            "(Iteration 7421 / 19600) loss: 1.190656\n",
            "(Iteration 7441 / 19600) loss: 1.041337\n",
            "(Iteration 7461 / 19600) loss: 1.398474\n",
            "(Iteration 7481 / 19600) loss: 1.120104\n",
            "(Iteration 7501 / 19600) loss: 1.243912\n",
            "(Iteration 7521 / 19600) loss: 1.230419\n",
            "(Iteration 7541 / 19600) loss: 1.051975\n",
            "(Iteration 7561 / 19600) loss: 1.134333\n",
            "(Iteration 7581 / 19600) loss: 1.040443\n",
            "(Iteration 7601 / 19600) loss: 0.973332\n",
            "(Iteration 7621 / 19600) loss: 0.919969\n",
            "(Iteration 7641 / 19600) loss: 1.112365\n",
            "(Iteration 7661 / 19600) loss: 1.283127\n",
            "(Iteration 7681 / 19600) loss: 1.087524\n",
            "(Iteration 7701 / 19600) loss: 0.884135\n",
            "(Iteration 7721 / 19600) loss: 1.100822\n",
            "(Iteration 7741 / 19600) loss: 1.014678\n",
            "(Iteration 7761 / 19600) loss: 1.307031\n",
            "(Iteration 7781 / 19600) loss: 1.243184\n",
            "(Iteration 7801 / 19600) loss: 1.054772\n",
            "(Iteration 7821 / 19600) loss: 0.868844\n",
            "(Epoch 8 / 20) train acc: 0.774000; val_acc: 0.705000\n",
            "(Iteration 7841 / 19600) loss: 1.047681\n",
            "(Iteration 7861 / 19600) loss: 1.647342\n",
            "(Iteration 7881 / 19600) loss: 1.493247\n",
            "(Iteration 7901 / 19600) loss: 1.262689\n",
            "(Iteration 7921 / 19600) loss: 1.120375\n",
            "(Iteration 7941 / 19600) loss: 1.263256\n",
            "(Iteration 7961 / 19600) loss: 0.993473\n",
            "(Iteration 7981 / 19600) loss: 1.375893\n",
            "(Iteration 8001 / 19600) loss: 1.168732\n",
            "(Iteration 8021 / 19600) loss: 1.132430\n",
            "(Iteration 8041 / 19600) loss: 1.094561\n",
            "(Iteration 8061 / 19600) loss: 1.221688\n",
            "(Iteration 8081 / 19600) loss: 1.187244\n",
            "(Iteration 8101 / 19600) loss: 0.990140\n",
            "(Iteration 8121 / 19600) loss: 1.240515\n",
            "(Iteration 8141 / 19600) loss: 1.038618\n",
            "(Iteration 8161 / 19600) loss: 1.161469\n",
            "(Iteration 8181 / 19600) loss: 1.282542\n",
            "(Iteration 8201 / 19600) loss: 1.199394\n",
            "(Iteration 8221 / 19600) loss: 1.038488\n",
            "(Iteration 8241 / 19600) loss: 1.271941\n",
            "(Iteration 8261 / 19600) loss: 1.012916\n",
            "(Iteration 8281 / 19600) loss: 1.386648\n",
            "(Iteration 8301 / 19600) loss: 1.028136\n",
            "(Iteration 8321 / 19600) loss: 1.301760\n",
            "(Iteration 8341 / 19600) loss: 1.211949\n",
            "(Iteration 8361 / 19600) loss: 1.322463\n",
            "(Iteration 8381 / 19600) loss: 0.939666\n",
            "(Iteration 8401 / 19600) loss: 1.427694\n",
            "(Iteration 8421 / 19600) loss: 1.227644\n",
            "(Iteration 8441 / 19600) loss: 1.192234\n",
            "(Iteration 8461 / 19600) loss: 1.102638\n",
            "(Iteration 8481 / 19600) loss: 1.301243\n",
            "(Iteration 8501 / 19600) loss: 1.350378\n",
            "(Iteration 8521 / 19600) loss: 1.196009\n",
            "(Iteration 8541 / 19600) loss: 0.988282\n",
            "(Iteration 8561 / 19600) loss: 1.369148\n",
            "(Iteration 8581 / 19600) loss: 1.357913\n",
            "(Iteration 8601 / 19600) loss: 1.242924\n",
            "(Iteration 8621 / 19600) loss: 1.115143\n",
            "(Iteration 8641 / 19600) loss: 1.022413\n",
            "(Iteration 8661 / 19600) loss: 1.164988\n",
            "(Iteration 8681 / 19600) loss: 1.144987\n",
            "(Iteration 8701 / 19600) loss: 1.085929\n",
            "(Iteration 8721 / 19600) loss: 1.016783\n",
            "(Iteration 8741 / 19600) loss: 1.176023\n",
            "(Iteration 8761 / 19600) loss: 0.899879\n",
            "(Iteration 8781 / 19600) loss: 1.384530\n",
            "(Iteration 8801 / 19600) loss: 1.084099\n",
            "(Epoch 9 / 20) train acc: 0.790000; val_acc: 0.705000\n",
            "(Iteration 8821 / 19600) loss: 1.101387\n",
            "(Iteration 8841 / 19600) loss: 1.077187\n",
            "(Iteration 8861 / 19600) loss: 1.311415\n",
            "(Iteration 8881 / 19600) loss: 1.325843\n",
            "(Iteration 8901 / 19600) loss: 1.288034\n",
            "(Iteration 8921 / 19600) loss: 1.244890\n",
            "(Iteration 8941 / 19600) loss: 1.194234\n",
            "(Iteration 8961 / 19600) loss: 1.269599\n",
            "(Iteration 8981 / 19600) loss: 1.023387\n",
            "(Iteration 9001 / 19600) loss: 1.395090\n",
            "(Iteration 9021 / 19600) loss: 1.331273\n",
            "(Iteration 9041 / 19600) loss: 1.346648\n",
            "(Iteration 9061 / 19600) loss: 1.324152\n",
            "(Iteration 9081 / 19600) loss: 1.147206\n",
            "(Iteration 9101 / 19600) loss: 1.103638\n",
            "(Iteration 9121 / 19600) loss: 1.389988\n",
            "(Iteration 9141 / 19600) loss: 1.172657\n",
            "(Iteration 9161 / 19600) loss: 1.327526\n",
            "(Iteration 9181 / 19600) loss: 1.382727\n",
            "(Iteration 9201 / 19600) loss: 1.021093\n",
            "(Iteration 9221 / 19600) loss: 1.097503\n",
            "(Iteration 9241 / 19600) loss: 1.200534\n",
            "(Iteration 9261 / 19600) loss: 1.524626\n",
            "(Iteration 9281 / 19600) loss: 1.093783\n",
            "(Iteration 9301 / 19600) loss: 1.382973\n",
            "(Iteration 9321 / 19600) loss: 1.121540\n",
            "(Iteration 9341 / 19600) loss: 1.194348\n",
            "(Iteration 9361 / 19600) loss: 1.046604\n",
            "(Iteration 9381 / 19600) loss: 1.293844\n",
            "(Iteration 9401 / 19600) loss: 1.079569\n",
            "(Iteration 9421 / 19600) loss: 0.927764\n",
            "(Iteration 9441 / 19600) loss: 1.054033\n",
            "(Iteration 9461 / 19600) loss: 1.170763\n",
            "(Iteration 9481 / 19600) loss: 1.369358\n",
            "(Iteration 9501 / 19600) loss: 1.139477\n",
            "(Iteration 9521 / 19600) loss: 1.421476\n",
            "(Iteration 9541 / 19600) loss: 1.411215\n",
            "(Iteration 9561 / 19600) loss: 1.233740\n",
            "(Iteration 9581 / 19600) loss: 1.067085\n",
            "(Iteration 9601 / 19600) loss: 1.167207\n",
            "(Iteration 9621 / 19600) loss: 1.210464\n",
            "(Iteration 9641 / 19600) loss: 1.315842\n",
            "(Iteration 9661 / 19600) loss: 1.035842\n",
            "(Iteration 9681 / 19600) loss: 1.104591\n",
            "(Iteration 9701 / 19600) loss: 1.170765\n",
            "(Iteration 9721 / 19600) loss: 1.258857\n",
            "(Iteration 9741 / 19600) loss: 1.275052\n",
            "(Iteration 9761 / 19600) loss: 1.052358\n",
            "(Iteration 9781 / 19600) loss: 0.998013\n",
            "(Epoch 10 / 20) train acc: 0.814000; val_acc: 0.703000\n",
            "(Iteration 9801 / 19600) loss: 1.074447\n",
            "(Iteration 9821 / 19600) loss: 1.060763\n",
            "(Iteration 9841 / 19600) loss: 1.345472\n",
            "(Iteration 9861 / 19600) loss: 1.401449\n",
            "(Iteration 9881 / 19600) loss: 1.222260\n",
            "(Iteration 9901 / 19600) loss: 1.234274\n",
            "(Iteration 9921 / 19600) loss: 1.148689\n",
            "(Iteration 9941 / 19600) loss: 1.000602\n",
            "(Iteration 9961 / 19600) loss: 1.481659\n",
            "(Iteration 9981 / 19600) loss: 0.994944\n",
            "(Iteration 10001 / 19600) loss: 1.284408\n",
            "(Iteration 10021 / 19600) loss: 0.864073\n",
            "(Iteration 10041 / 19600) loss: 1.077070\n",
            "(Iteration 10061 / 19600) loss: 1.078190\n",
            "(Iteration 10081 / 19600) loss: 1.068249\n",
            "(Iteration 10101 / 19600) loss: 1.127797\n",
            "(Iteration 10121 / 19600) loss: 1.192930\n",
            "(Iteration 10141 / 19600) loss: 1.155361\n",
            "(Iteration 10161 / 19600) loss: 1.318434\n",
            "(Iteration 10181 / 19600) loss: 1.279047\n",
            "(Iteration 10201 / 19600) loss: 1.167428\n",
            "(Iteration 10221 / 19600) loss: 1.155653\n",
            "(Iteration 10241 / 19600) loss: 1.285752\n",
            "(Iteration 10261 / 19600) loss: 0.829092\n",
            "(Iteration 10281 / 19600) loss: 0.981733\n",
            "(Iteration 10301 / 19600) loss: 0.932622\n",
            "(Iteration 10321 / 19600) loss: 1.208741\n",
            "(Iteration 10341 / 19600) loss: 1.040813\n",
            "(Iteration 10361 / 19600) loss: 1.410078\n",
            "(Iteration 10381 / 19600) loss: 1.053504\n",
            "(Iteration 10401 / 19600) loss: 1.072710\n",
            "(Iteration 10421 / 19600) loss: 1.249301\n",
            "(Iteration 10441 / 19600) loss: 0.898808\n",
            "(Iteration 10461 / 19600) loss: 1.175707\n",
            "(Iteration 10481 / 19600) loss: 1.207167\n",
            "(Iteration 10501 / 19600) loss: 1.104559\n",
            "(Iteration 10521 / 19600) loss: 1.035628\n",
            "(Iteration 10541 / 19600) loss: 1.106745\n",
            "(Iteration 10561 / 19600) loss: 1.017779\n",
            "(Iteration 10581 / 19600) loss: 1.204917\n",
            "(Iteration 10601 / 19600) loss: 1.303841\n",
            "(Iteration 10621 / 19600) loss: 0.949680\n",
            "(Iteration 10641 / 19600) loss: 0.992239\n",
            "(Iteration 10661 / 19600) loss: 1.467082\n",
            "(Iteration 10681 / 19600) loss: 1.052297\n",
            "(Iteration 10701 / 19600) loss: 0.991532\n",
            "(Iteration 10721 / 19600) loss: 1.268773\n",
            "(Iteration 10741 / 19600) loss: 0.996821\n",
            "(Iteration 10761 / 19600) loss: 1.213245\n",
            "(Epoch 11 / 20) train acc: 0.833000; val_acc: 0.736000\n",
            "(Iteration 10781 / 19600) loss: 1.012893\n",
            "(Iteration 10801 / 19600) loss: 1.107827\n",
            "(Iteration 10821 / 19600) loss: 1.168875\n",
            "(Iteration 10841 / 19600) loss: 1.084699\n",
            "(Iteration 10861 / 19600) loss: 1.035060\n",
            "(Iteration 10881 / 19600) loss: 1.227040\n",
            "(Iteration 10901 / 19600) loss: 0.953278\n",
            "(Iteration 10921 / 19600) loss: 1.094046\n",
            "(Iteration 10941 / 19600) loss: 1.039537\n",
            "(Iteration 10961 / 19600) loss: 1.255925\n",
            "(Iteration 10981 / 19600) loss: 1.280110\n",
            "(Iteration 11001 / 19600) loss: 1.176254\n",
            "(Iteration 11021 / 19600) loss: 0.980005\n",
            "(Iteration 11041 / 19600) loss: 0.968378\n",
            "(Iteration 11061 / 19600) loss: 1.451034\n",
            "(Iteration 11081 / 19600) loss: 1.176890\n",
            "(Iteration 11101 / 19600) loss: 1.450939\n",
            "(Iteration 11121 / 19600) loss: 1.341480\n",
            "(Iteration 11141 / 19600) loss: 1.175899\n",
            "(Iteration 11161 / 19600) loss: 1.159937\n",
            "(Iteration 11181 / 19600) loss: 1.081046\n",
            "(Iteration 11201 / 19600) loss: 1.105910\n",
            "(Iteration 11221 / 19600) loss: 0.980516\n",
            "(Iteration 11241 / 19600) loss: 0.918665\n",
            "(Iteration 11261 / 19600) loss: 1.335652\n",
            "(Iteration 11281 / 19600) loss: 1.065216\n",
            "(Iteration 11301 / 19600) loss: 1.061810\n",
            "(Iteration 11321 / 19600) loss: 1.075984\n",
            "(Iteration 11341 / 19600) loss: 1.312477\n",
            "(Iteration 11361 / 19600) loss: 1.190052\n",
            "(Iteration 11381 / 19600) loss: 1.190607\n",
            "(Iteration 11401 / 19600) loss: 1.044879\n",
            "(Iteration 11421 / 19600) loss: 1.310536\n",
            "(Iteration 11441 / 19600) loss: 1.123507\n",
            "(Iteration 11461 / 19600) loss: 1.164524\n",
            "(Iteration 11481 / 19600) loss: 1.247241\n",
            "(Iteration 11501 / 19600) loss: 1.296045\n",
            "(Iteration 11521 / 19600) loss: 0.875170\n",
            "(Iteration 11541 / 19600) loss: 0.896803\n",
            "(Iteration 11561 / 19600) loss: 1.057353\n",
            "(Iteration 11581 / 19600) loss: 1.228638\n",
            "(Iteration 11601 / 19600) loss: 1.132930\n",
            "(Iteration 11621 / 19600) loss: 1.164570\n",
            "(Iteration 11641 / 19600) loss: 1.058944\n",
            "(Iteration 11661 / 19600) loss: 1.367837\n",
            "(Iteration 11681 / 19600) loss: 1.323216\n",
            "(Iteration 11701 / 19600) loss: 1.195799\n",
            "(Iteration 11721 / 19600) loss: 1.043742\n",
            "(Iteration 11741 / 19600) loss: 1.169128\n",
            "(Epoch 12 / 20) train acc: 0.831000; val_acc: 0.725000\n",
            "(Iteration 11761 / 19600) loss: 1.179095\n",
            "(Iteration 11781 / 19600) loss: 1.438285\n",
            "(Iteration 11801 / 19600) loss: 1.155654\n",
            "(Iteration 11821 / 19600) loss: 1.085205\n",
            "(Iteration 11841 / 19600) loss: 0.998962\n",
            "(Iteration 11861 / 19600) loss: 1.294680\n",
            "(Iteration 11881 / 19600) loss: 1.072351\n",
            "(Iteration 11901 / 19600) loss: 1.170226\n",
            "(Iteration 11921 / 19600) loss: 1.055632\n",
            "(Iteration 11941 / 19600) loss: 1.353936\n",
            "(Iteration 11961 / 19600) loss: 1.122245\n",
            "(Iteration 11981 / 19600) loss: 0.965216\n",
            "(Iteration 12001 / 19600) loss: 1.501886\n",
            "(Iteration 12021 / 19600) loss: 1.219835\n",
            "(Iteration 12041 / 19600) loss: 1.039963\n",
            "(Iteration 12061 / 19600) loss: 1.200219\n",
            "(Iteration 12081 / 19600) loss: 1.292554\n",
            "(Iteration 12101 / 19600) loss: 1.149044\n",
            "(Iteration 12121 / 19600) loss: 1.349996\n",
            "(Iteration 12141 / 19600) loss: 1.042081\n",
            "(Iteration 12161 / 19600) loss: 0.895921\n",
            "(Iteration 12181 / 19600) loss: 0.924759\n",
            "(Iteration 12201 / 19600) loss: 1.118280\n",
            "(Iteration 12221 / 19600) loss: 1.066912\n",
            "(Iteration 12241 / 19600) loss: 1.128661\n",
            "(Iteration 12261 / 19600) loss: 0.868221\n",
            "(Iteration 12281 / 19600) loss: 1.132993\n",
            "(Iteration 12301 / 19600) loss: 1.158955\n",
            "(Iteration 12321 / 19600) loss: 1.063544\n",
            "(Iteration 12341 / 19600) loss: 1.344590\n",
            "(Iteration 12361 / 19600) loss: 1.014278\n",
            "(Iteration 12381 / 19600) loss: 1.168552\n",
            "(Iteration 12401 / 19600) loss: 1.052308\n",
            "(Iteration 12421 / 19600) loss: 1.136287\n",
            "(Iteration 12441 / 19600) loss: 1.219572\n",
            "(Iteration 12461 / 19600) loss: 1.009202\n",
            "(Iteration 12481 / 19600) loss: 1.130449\n",
            "(Iteration 12501 / 19600) loss: 1.033042\n",
            "(Iteration 12521 / 19600) loss: 0.995290\n",
            "(Iteration 12541 / 19600) loss: 1.274776\n",
            "(Iteration 12561 / 19600) loss: 1.158328\n",
            "(Iteration 12581 / 19600) loss: 0.995219\n",
            "(Iteration 12601 / 19600) loss: 1.128141\n",
            "(Iteration 12621 / 19600) loss: 1.183352\n",
            "(Iteration 12641 / 19600) loss: 1.449877\n",
            "(Iteration 12661 / 19600) loss: 1.369623\n",
            "(Iteration 12681 / 19600) loss: 1.018307\n",
            "(Iteration 12701 / 19600) loss: 1.125454\n",
            "(Iteration 12721 / 19600) loss: 1.057966\n",
            "(Epoch 13 / 20) train acc: 0.832000; val_acc: 0.724000\n",
            "(Iteration 12741 / 19600) loss: 1.078742\n",
            "(Iteration 12761 / 19600) loss: 1.385869\n",
            "(Iteration 12781 / 19600) loss: 1.176723\n",
            "(Iteration 12801 / 19600) loss: 1.049852\n",
            "(Iteration 12821 / 19600) loss: 1.022826\n",
            "(Iteration 12841 / 19600) loss: 1.218821\n",
            "(Iteration 12861 / 19600) loss: 1.217484\n",
            "(Iteration 12881 / 19600) loss: 1.046611\n",
            "(Iteration 12901 / 19600) loss: 1.103164\n",
            "(Iteration 12921 / 19600) loss: 1.080483\n",
            "(Iteration 12941 / 19600) loss: 1.136937\n",
            "(Iteration 12961 / 19600) loss: 1.063683\n",
            "(Iteration 12981 / 19600) loss: 1.122171\n",
            "(Iteration 13001 / 19600) loss: 1.326379\n",
            "(Iteration 13021 / 19600) loss: 1.171194\n",
            "(Iteration 13041 / 19600) loss: 1.061366\n",
            "(Iteration 13061 / 19600) loss: 1.121320\n",
            "(Iteration 13081 / 19600) loss: 1.030170\n",
            "(Iteration 13101 / 19600) loss: 0.826251\n",
            "(Iteration 13121 / 19600) loss: 0.932609\n",
            "(Iteration 13141 / 19600) loss: 1.322100\n",
            "(Iteration 13161 / 19600) loss: 1.508044\n",
            "(Iteration 13181 / 19600) loss: 0.950027\n",
            "(Iteration 13201 / 19600) loss: 1.201947\n",
            "(Iteration 13221 / 19600) loss: 1.199365\n",
            "(Iteration 13241 / 19600) loss: 1.058966\n",
            "(Iteration 13261 / 19600) loss: 0.873732\n",
            "(Iteration 13281 / 19600) loss: 1.168039\n",
            "(Iteration 13301 / 19600) loss: 1.173541\n",
            "(Iteration 13321 / 19600) loss: 1.060543\n",
            "(Iteration 13341 / 19600) loss: 1.419887\n",
            "(Iteration 13361 / 19600) loss: 1.014190\n",
            "(Iteration 13381 / 19600) loss: 1.339850\n",
            "(Iteration 13401 / 19600) loss: 1.493709\n",
            "(Iteration 13421 / 19600) loss: 1.173926\n",
            "(Iteration 13441 / 19600) loss: 1.289374\n",
            "(Iteration 13461 / 19600) loss: 0.821396\n",
            "(Iteration 13481 / 19600) loss: 1.152523\n",
            "(Iteration 13501 / 19600) loss: 1.258889\n",
            "(Iteration 13521 / 19600) loss: 1.088111\n",
            "(Iteration 13541 / 19600) loss: 0.991127\n",
            "(Iteration 13561 / 19600) loss: 1.214735\n",
            "(Iteration 13581 / 19600) loss: 1.225758\n",
            "(Iteration 13601 / 19600) loss: 1.113477\n",
            "(Iteration 13621 / 19600) loss: 1.025006\n",
            "(Iteration 13641 / 19600) loss: 1.134323\n",
            "(Iteration 13661 / 19600) loss: 1.323647\n",
            "(Iteration 13681 / 19600) loss: 0.876699\n",
            "(Iteration 13701 / 19600) loss: 1.007386\n",
            "(Epoch 14 / 20) train acc: 0.843000; val_acc: 0.718000\n",
            "(Iteration 13721 / 19600) loss: 1.100695\n",
            "(Iteration 13741 / 19600) loss: 1.111382\n",
            "(Iteration 13761 / 19600) loss: 1.269243\n",
            "(Iteration 13781 / 19600) loss: 1.053806\n",
            "(Iteration 13801 / 19600) loss: 1.201857\n",
            "(Iteration 13821 / 19600) loss: 1.128086\n",
            "(Iteration 13841 / 19600) loss: 1.180944\n",
            "(Iteration 13861 / 19600) loss: 1.281263\n",
            "(Iteration 13881 / 19600) loss: 1.242220\n",
            "(Iteration 13901 / 19600) loss: 1.363207\n",
            "(Iteration 13921 / 19600) loss: 1.330950\n",
            "(Iteration 13941 / 19600) loss: 1.041398\n",
            "(Iteration 13961 / 19600) loss: 1.235270\n",
            "(Iteration 13981 / 19600) loss: 0.996967\n",
            "(Iteration 14001 / 19600) loss: 1.134020\n",
            "(Iteration 14021 / 19600) loss: 1.104884\n",
            "(Iteration 14041 / 19600) loss: 0.957873\n",
            "(Iteration 14061 / 19600) loss: 0.921799\n",
            "(Iteration 14081 / 19600) loss: 1.395369\n",
            "(Iteration 14101 / 19600) loss: 0.937313\n",
            "(Iteration 14121 / 19600) loss: 1.026749\n",
            "(Iteration 14141 / 19600) loss: 1.061953\n",
            "(Iteration 14161 / 19600) loss: 1.159322\n",
            "(Iteration 14181 / 19600) loss: 1.022081\n",
            "(Iteration 14201 / 19600) loss: 1.214323\n",
            "(Iteration 14221 / 19600) loss: 1.003050\n",
            "(Iteration 14241 / 19600) loss: 1.083604\n",
            "(Iteration 14261 / 19600) loss: 1.008253\n",
            "(Iteration 14281 / 19600) loss: 1.223707\n",
            "(Iteration 14301 / 19600) loss: 1.170315\n",
            "(Iteration 14321 / 19600) loss: 1.246896\n",
            "(Iteration 14341 / 19600) loss: 1.208981\n",
            "(Iteration 14361 / 19600) loss: 1.269470\n",
            "(Iteration 14381 / 19600) loss: 1.274429\n",
            "(Iteration 14401 / 19600) loss: 0.906011\n",
            "(Iteration 14421 / 19600) loss: 1.166194\n",
            "(Iteration 14441 / 19600) loss: 1.064196\n",
            "(Iteration 14461 / 19600) loss: 1.099680\n",
            "(Iteration 14481 / 19600) loss: 0.887847\n",
            "(Iteration 14501 / 19600) loss: 0.967931\n",
            "(Iteration 14521 / 19600) loss: 1.050615\n",
            "(Iteration 14541 / 19600) loss: 1.293724\n",
            "(Iteration 14561 / 19600) loss: 1.145259\n",
            "(Iteration 14581 / 19600) loss: 1.075966\n",
            "(Iteration 14601 / 19600) loss: 1.363525\n",
            "(Iteration 14621 / 19600) loss: 1.031074\n",
            "(Iteration 14641 / 19600) loss: 0.939619\n",
            "(Iteration 14661 / 19600) loss: 1.362793\n",
            "(Iteration 14681 / 19600) loss: 1.151987\n",
            "(Epoch 15 / 20) train acc: 0.812000; val_acc: 0.713000\n",
            "(Iteration 14701 / 19600) loss: 1.268027\n",
            "(Iteration 14721 / 19600) loss: 1.313705\n",
            "(Iteration 14741 / 19600) loss: 1.440701\n",
            "(Iteration 14761 / 19600) loss: 1.352938\n",
            "(Iteration 14781 / 19600) loss: 1.015588\n",
            "(Iteration 14801 / 19600) loss: 1.081321\n",
            "(Iteration 14821 / 19600) loss: 1.110934\n",
            "(Iteration 14841 / 19600) loss: 1.085602\n",
            "(Iteration 14861 / 19600) loss: 1.060654\n",
            "(Iteration 14881 / 19600) loss: 1.074596\n",
            "(Iteration 14901 / 19600) loss: 1.219030\n",
            "(Iteration 14921 / 19600) loss: 1.024525\n",
            "(Iteration 14941 / 19600) loss: 1.235370\n",
            "(Iteration 14961 / 19600) loss: 1.208391\n",
            "(Iteration 14981 / 19600) loss: 1.036635\n",
            "(Iteration 15001 / 19600) loss: 1.019677\n",
            "(Iteration 15021 / 19600) loss: 1.157251\n",
            "(Iteration 15041 / 19600) loss: 1.282499\n",
            "(Iteration 15061 / 19600) loss: 1.192848\n",
            "(Iteration 15081 / 19600) loss: 0.989960\n",
            "(Iteration 15101 / 19600) loss: 1.126499\n",
            "(Iteration 15121 / 19600) loss: 1.013060\n",
            "(Iteration 15141 / 19600) loss: 0.969923\n",
            "(Iteration 15161 / 19600) loss: 0.991134\n",
            "(Iteration 15181 / 19600) loss: 1.071667\n",
            "(Iteration 15201 / 19600) loss: 1.143084\n",
            "(Iteration 15221 / 19600) loss: 1.177939\n",
            "(Iteration 15241 / 19600) loss: 1.113503\n",
            "(Iteration 15261 / 19600) loss: 1.132278\n",
            "(Iteration 15281 / 19600) loss: 1.162164\n",
            "(Iteration 15301 / 19600) loss: 1.202525\n",
            "(Iteration 15321 / 19600) loss: 1.060677\n",
            "(Iteration 15341 / 19600) loss: 0.984132\n",
            "(Iteration 15361 / 19600) loss: 1.176239\n",
            "(Iteration 15381 / 19600) loss: 1.251536\n",
            "(Iteration 15401 / 19600) loss: 1.086449\n",
            "(Iteration 15421 / 19600) loss: 1.093798\n",
            "(Iteration 15441 / 19600) loss: 1.272351\n",
            "(Iteration 15461 / 19600) loss: 1.173408\n",
            "(Iteration 15481 / 19600) loss: 1.190752\n",
            "(Iteration 15501 / 19600) loss: 1.226475\n",
            "(Iteration 15521 / 19600) loss: 1.332120\n",
            "(Iteration 15541 / 19600) loss: 0.914150\n",
            "(Iteration 15561 / 19600) loss: 1.300698\n",
            "(Iteration 15581 / 19600) loss: 1.328383\n",
            "(Iteration 15601 / 19600) loss: 0.986403\n",
            "(Iteration 15621 / 19600) loss: 1.213172\n",
            "(Iteration 15641 / 19600) loss: 1.062562\n",
            "(Iteration 15661 / 19600) loss: 1.147429\n",
            "(Epoch 16 / 20) train acc: 0.833000; val_acc: 0.736000\n",
            "(Iteration 15681 / 19600) loss: 1.038872\n",
            "(Iteration 15701 / 19600) loss: 1.336630\n",
            "(Iteration 15721 / 19600) loss: 1.166776\n",
            "(Iteration 15741 / 19600) loss: 1.005060\n",
            "(Iteration 15761 / 19600) loss: 1.210746\n",
            "(Iteration 15781 / 19600) loss: 0.993009\n",
            "(Iteration 15801 / 19600) loss: 1.105459\n",
            "(Iteration 15821 / 19600) loss: 1.138092\n",
            "(Iteration 15841 / 19600) loss: 0.983471\n",
            "(Iteration 15861 / 19600) loss: 1.191528\n",
            "(Iteration 15881 / 19600) loss: 1.273222\n",
            "(Iteration 15901 / 19600) loss: 1.157038\n",
            "(Iteration 15921 / 19600) loss: 1.282804\n",
            "(Iteration 15941 / 19600) loss: 1.026314\n",
            "(Iteration 15961 / 19600) loss: 1.044799\n",
            "(Iteration 15981 / 19600) loss: 1.115152\n",
            "(Iteration 16001 / 19600) loss: 1.312352\n",
            "(Iteration 16021 / 19600) loss: 1.061284\n",
            "(Iteration 16041 / 19600) loss: 1.065541\n",
            "(Iteration 16061 / 19600) loss: 1.055336\n",
            "(Iteration 16081 / 19600) loss: 1.013392\n",
            "(Iteration 16101 / 19600) loss: 1.057118\n",
            "(Iteration 16121 / 19600) loss: 0.943036\n",
            "(Iteration 16141 / 19600) loss: 1.121741\n",
            "(Iteration 16161 / 19600) loss: 1.110318\n",
            "(Iteration 16181 / 19600) loss: 0.968561\n",
            "(Iteration 16201 / 19600) loss: 1.108612\n",
            "(Iteration 16221 / 19600) loss: 1.040112\n",
            "(Iteration 16241 / 19600) loss: 0.896394\n",
            "(Iteration 16261 / 19600) loss: 1.160393\n",
            "(Iteration 16281 / 19600) loss: 1.253490\n",
            "(Iteration 16301 / 19600) loss: 1.089825\n",
            "(Iteration 16321 / 19600) loss: 1.359619\n",
            "(Iteration 16341 / 19600) loss: 0.907201\n",
            "(Iteration 16361 / 19600) loss: 1.278212\n",
            "(Iteration 16381 / 19600) loss: 1.127150\n",
            "(Iteration 16401 / 19600) loss: 1.031088\n",
            "(Iteration 16421 / 19600) loss: 0.781590\n",
            "(Iteration 16441 / 19600) loss: 1.074705\n",
            "(Iteration 16461 / 19600) loss: 1.362588\n",
            "(Iteration 16481 / 19600) loss: 1.429623\n",
            "(Iteration 16501 / 19600) loss: 1.289906\n",
            "(Iteration 16521 / 19600) loss: 1.220504\n",
            "(Iteration 16541 / 19600) loss: 1.128503\n",
            "(Iteration 16561 / 19600) loss: 1.234894\n",
            "(Iteration 16581 / 19600) loss: 1.043282\n",
            "(Iteration 16601 / 19600) loss: 1.128396\n",
            "(Iteration 16621 / 19600) loss: 1.071573\n",
            "(Iteration 16641 / 19600) loss: 1.232782\n",
            "(Epoch 17 / 20) train acc: 0.847000; val_acc: 0.726000\n",
            "(Iteration 16661 / 19600) loss: 1.075409\n",
            "(Iteration 16681 / 19600) loss: 1.187735\n",
            "(Iteration 16701 / 19600) loss: 1.166564\n",
            "(Iteration 16721 / 19600) loss: 1.144528\n",
            "(Iteration 16741 / 19600) loss: 1.150285\n",
            "(Iteration 16761 / 19600) loss: 1.200067\n",
            "(Iteration 16781 / 19600) loss: 1.026392\n",
            "(Iteration 16801 / 19600) loss: 0.932657\n",
            "(Iteration 16821 / 19600) loss: 1.247685\n",
            "(Iteration 16841 / 19600) loss: 1.022509\n",
            "(Iteration 16861 / 19600) loss: 1.340145\n",
            "(Iteration 16881 / 19600) loss: 1.059579\n",
            "(Iteration 16901 / 19600) loss: 1.214087\n",
            "(Iteration 16921 / 19600) loss: 1.207458\n",
            "(Iteration 16941 / 19600) loss: 0.918652\n",
            "(Iteration 16961 / 19600) loss: 1.094807\n",
            "(Iteration 16981 / 19600) loss: 1.067382\n",
            "(Iteration 17001 / 19600) loss: 1.308637\n",
            "(Iteration 17021 / 19600) loss: 1.299420\n",
            "(Iteration 17041 / 19600) loss: 1.220100\n",
            "(Iteration 17061 / 19600) loss: 1.360514\n",
            "(Iteration 17081 / 19600) loss: 1.292465\n",
            "(Iteration 17101 / 19600) loss: 1.155935\n",
            "(Iteration 17121 / 19600) loss: 1.013005\n",
            "(Iteration 17141 / 19600) loss: 0.855624\n",
            "(Iteration 17161 / 19600) loss: 1.161459\n",
            "(Iteration 17181 / 19600) loss: 1.097382\n",
            "(Iteration 17201 / 19600) loss: 1.200638\n",
            "(Iteration 17221 / 19600) loss: 1.188082\n",
            "(Iteration 17241 / 19600) loss: 0.937977\n",
            "(Iteration 17261 / 19600) loss: 1.223297\n",
            "(Iteration 17281 / 19600) loss: 1.106059\n",
            "(Iteration 17301 / 19600) loss: 0.999001\n",
            "(Iteration 17321 / 19600) loss: 1.389757\n",
            "(Iteration 17341 / 19600) loss: 1.080068\n",
            "(Iteration 17361 / 19600) loss: 0.981649\n",
            "(Iteration 17381 / 19600) loss: 1.133689\n",
            "(Iteration 17401 / 19600) loss: 0.944792\n",
            "(Iteration 17421 / 19600) loss: 1.238670\n",
            "(Iteration 17441 / 19600) loss: 0.991626\n",
            "(Iteration 17461 / 19600) loss: 0.866975\n",
            "(Iteration 17481 / 19600) loss: 1.001387\n",
            "(Iteration 17501 / 19600) loss: 0.930245\n",
            "(Iteration 17521 / 19600) loss: 0.981696\n",
            "(Iteration 17541 / 19600) loss: 1.168697\n",
            "(Iteration 17561 / 19600) loss: 1.003617\n",
            "(Iteration 17581 / 19600) loss: 1.420110\n",
            "(Iteration 17601 / 19600) loss: 1.171744\n",
            "(Iteration 17621 / 19600) loss: 1.184385\n",
            "(Epoch 18 / 20) train acc: 0.845000; val_acc: 0.744000\n",
            "(Iteration 17641 / 19600) loss: 0.830135\n",
            "(Iteration 17661 / 19600) loss: 1.231432\n",
            "(Iteration 17681 / 19600) loss: 1.196325\n",
            "(Iteration 17701 / 19600) loss: 1.259141\n",
            "(Iteration 17721 / 19600) loss: 1.073114\n",
            "(Iteration 17741 / 19600) loss: 1.120717\n",
            "(Iteration 17761 / 19600) loss: 0.977075\n",
            "(Iteration 17781 / 19600) loss: 1.145377\n",
            "(Iteration 17801 / 19600) loss: 1.172362\n",
            "(Iteration 17821 / 19600) loss: 1.135907\n",
            "(Iteration 17841 / 19600) loss: 1.234113\n",
            "(Iteration 17861 / 19600) loss: 1.224296\n",
            "(Iteration 17881 / 19600) loss: 1.141130\n",
            "(Iteration 17901 / 19600) loss: 0.980982\n",
            "(Iteration 17921 / 19600) loss: 1.286765\n",
            "(Iteration 17941 / 19600) loss: 1.189072\n",
            "(Iteration 17961 / 19600) loss: 0.975374\n",
            "(Iteration 17981 / 19600) loss: 1.124141\n",
            "(Iteration 18001 / 19600) loss: 1.035648\n",
            "(Iteration 18021 / 19600) loss: 0.937720\n",
            "(Iteration 18041 / 19600) loss: 1.276224\n",
            "(Iteration 18061 / 19600) loss: 1.256351\n",
            "(Iteration 18081 / 19600) loss: 1.056865\n",
            "(Iteration 18101 / 19600) loss: 1.342791\n",
            "(Iteration 18121 / 19600) loss: 1.417333\n",
            "(Iteration 18141 / 19600) loss: 1.043951\n",
            "(Iteration 18161 / 19600) loss: 1.079000\n",
            "(Iteration 18181 / 19600) loss: 1.008863\n",
            "(Iteration 18201 / 19600) loss: 1.045588\n",
            "(Iteration 18221 / 19600) loss: 1.064378\n",
            "(Iteration 18241 / 19600) loss: 1.034348\n",
            "(Iteration 18261 / 19600) loss: 1.193678\n",
            "(Iteration 18281 / 19600) loss: 1.112803\n",
            "(Iteration 18301 / 19600) loss: 1.322521\n",
            "(Iteration 18321 / 19600) loss: 1.212197\n",
            "(Iteration 18341 / 19600) loss: 1.159151\n",
            "(Iteration 18361 / 19600) loss: 1.032064\n",
            "(Iteration 18381 / 19600) loss: 1.220686\n",
            "(Iteration 18401 / 19600) loss: 1.122736\n",
            "(Iteration 18421 / 19600) loss: 0.895194\n",
            "(Iteration 18441 / 19600) loss: 1.133482\n",
            "(Iteration 18461 / 19600) loss: 1.063667\n",
            "(Iteration 18481 / 19600) loss: 1.142683\n",
            "(Iteration 18501 / 19600) loss: 1.004569\n",
            "(Iteration 18521 / 19600) loss: 1.005415\n",
            "(Iteration 18541 / 19600) loss: 1.233941\n",
            "(Iteration 18561 / 19600) loss: 1.081391\n",
            "(Iteration 18581 / 19600) loss: 1.214328\n",
            "(Iteration 18601 / 19600) loss: 1.047335\n",
            "(Epoch 19 / 20) train acc: 0.841000; val_acc: 0.734000\n",
            "(Iteration 18621 / 19600) loss: 1.224982\n",
            "(Iteration 18641 / 19600) loss: 1.000832\n",
            "(Iteration 18661 / 19600) loss: 1.060971\n",
            "(Iteration 18681 / 19600) loss: 0.909628\n",
            "(Iteration 18701 / 19600) loss: 1.106426\n",
            "(Iteration 18721 / 19600) loss: 1.073603\n",
            "(Iteration 18741 / 19600) loss: 1.282789\n",
            "(Iteration 18761 / 19600) loss: 0.972075\n",
            "(Iteration 18781 / 19600) loss: 1.184743\n",
            "(Iteration 18801 / 19600) loss: 0.993871\n",
            "(Iteration 18821 / 19600) loss: 0.985621\n",
            "(Iteration 18841 / 19600) loss: 1.008023\n",
            "(Iteration 18861 / 19600) loss: 1.050021\n",
            "(Iteration 18881 / 19600) loss: 0.844629\n",
            "(Iteration 18901 / 19600) loss: 1.099659\n",
            "(Iteration 18921 / 19600) loss: 1.109706\n",
            "(Iteration 18941 / 19600) loss: 0.944462\n",
            "(Iteration 18961 / 19600) loss: 0.979572\n",
            "(Iteration 18981 / 19600) loss: 0.896260\n",
            "(Iteration 19001 / 19600) loss: 0.877044\n",
            "(Iteration 19021 / 19600) loss: 1.015631\n",
            "(Iteration 19041 / 19600) loss: 1.281398\n",
            "(Iteration 19061 / 19600) loss: 1.359094\n",
            "(Iteration 19081 / 19600) loss: 1.025064\n",
            "(Iteration 19101 / 19600) loss: 1.013189\n",
            "(Iteration 19121 / 19600) loss: 0.997182\n",
            "(Iteration 19141 / 19600) loss: 1.203713\n",
            "(Iteration 19161 / 19600) loss: 1.153245\n",
            "(Iteration 19181 / 19600) loss: 0.973718\n",
            "(Iteration 19201 / 19600) loss: 1.381436\n",
            "(Iteration 19221 / 19600) loss: 1.080936\n",
            "(Iteration 19241 / 19600) loss: 1.135512\n",
            "(Iteration 19261 / 19600) loss: 1.246778\n",
            "(Iteration 19281 / 19600) loss: 1.259275\n",
            "(Iteration 19301 / 19600) loss: 1.069925\n",
            "(Iteration 19321 / 19600) loss: 1.049934\n",
            "(Iteration 19341 / 19600) loss: 0.903948\n",
            "(Iteration 19361 / 19600) loss: 0.996182\n",
            "(Iteration 19381 / 19600) loss: 0.864767\n",
            "(Iteration 19401 / 19600) loss: 0.926886\n",
            "(Iteration 19421 / 19600) loss: 1.342741\n",
            "(Iteration 19441 / 19600) loss: 0.950681\n",
            "(Iteration 19461 / 19600) loss: 1.097456\n",
            "(Iteration 19481 / 19600) loss: 1.051275\n",
            "(Iteration 19501 / 19600) loss: 1.113620\n",
            "(Iteration 19521 / 19600) loss: 1.103354\n",
            "(Iteration 19541 / 19600) loss: 1.253298\n",
            "(Iteration 19561 / 19600) loss: 0.987436\n",
            "(Iteration 19581 / 19600) loss: 1.099023\n",
            "(Epoch 20 / 20) train acc: 0.849000; val_acc: 0.722000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ciq7rzyWKSBC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 8534
        },
        "outputId": "1abdffd5-450a-4bda-f4a3-cd51ddda671a"
      },
      "cell_type": "code",
      "source": [
        "# model = MultiLayerConvNet(weight_scale=0.001, hidden_dim=[800,400,200], reg=0.001, dropout = 0.5)\n",
        "\n",
        "# solver = Solver(model, data,\n",
        "#                 num_epochs=20, batch_size=50,\n",
        "#                 update_rule='adam',\n",
        "#                 optim_config={\n",
        "#                   'learning_rate': 1e-3,\n",
        "#                 },\n",
        "#                 verbose=True, print_every=20)\n",
        "# solver.train()"
      ],
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(Iteration 1 / 9800) loss: 2.307273\n",
            "(Epoch 0 / 10) train acc: 0.099000; val_acc: 0.101000\n",
            "(Iteration 21 / 9800) loss: 2.079409\n",
            "(Iteration 41 / 9800) loss: 1.835074\n",
            "(Iteration 61 / 9800) loss: 1.793039\n",
            "(Iteration 81 / 9800) loss: 1.528854\n",
            "(Iteration 101 / 9800) loss: 1.624514\n",
            "(Iteration 121 / 9800) loss: 1.605157\n",
            "(Iteration 141 / 9800) loss: 1.785764\n",
            "(Iteration 161 / 9800) loss: 1.532722\n",
            "(Iteration 181 / 9800) loss: 1.745724\n",
            "(Iteration 201 / 9800) loss: 1.798177\n",
            "(Iteration 221 / 9800) loss: 1.664266\n",
            "(Iteration 241 / 9800) loss: 1.554599\n",
            "(Iteration 261 / 9800) loss: 1.659986\n",
            "(Iteration 281 / 9800) loss: 1.607672\n",
            "(Iteration 301 / 9800) loss: 1.262297\n",
            "(Iteration 321 / 9800) loss: 1.433269\n",
            "(Iteration 341 / 9800) loss: 1.469270\n",
            "(Iteration 361 / 9800) loss: 1.698848\n",
            "(Iteration 381 / 9800) loss: 1.465176\n",
            "(Iteration 401 / 9800) loss: 1.490384\n",
            "(Iteration 421 / 9800) loss: 1.219396\n",
            "(Iteration 441 / 9800) loss: 1.450579\n",
            "(Iteration 461 / 9800) loss: 1.334288\n",
            "(Iteration 481 / 9800) loss: 1.384253\n",
            "(Iteration 501 / 9800) loss: 1.435014\n",
            "(Iteration 521 / 9800) loss: 1.330154\n",
            "(Iteration 541 / 9800) loss: 1.471881\n",
            "(Iteration 561 / 9800) loss: 1.438666\n",
            "(Iteration 581 / 9800) loss: 1.282103\n",
            "(Iteration 601 / 9800) loss: 1.582022\n",
            "(Iteration 621 / 9800) loss: 1.624105\n",
            "(Iteration 641 / 9800) loss: 1.457041\n",
            "(Iteration 661 / 9800) loss: 1.860271\n",
            "(Iteration 681 / 9800) loss: 1.423878\n",
            "(Iteration 701 / 9800) loss: 1.495193\n",
            "(Iteration 721 / 9800) loss: 1.346626\n",
            "(Iteration 741 / 9800) loss: 1.305482\n",
            "(Iteration 761 / 9800) loss: 1.947309\n",
            "(Iteration 781 / 9800) loss: 1.700867\n",
            "(Iteration 801 / 9800) loss: 1.432151\n",
            "(Iteration 821 / 9800) loss: 1.389913\n",
            "(Iteration 841 / 9800) loss: 1.369684\n",
            "(Iteration 861 / 9800) loss: 1.279333\n",
            "(Iteration 881 / 9800) loss: 1.436539\n",
            "(Iteration 901 / 9800) loss: 1.444999\n",
            "(Iteration 921 / 9800) loss: 1.392601\n",
            "(Iteration 941 / 9800) loss: 1.499157\n",
            "(Iteration 961 / 9800) loss: 1.345764\n",
            "(Epoch 1 / 10) train acc: 0.618000; val_acc: 0.635000\n",
            "(Iteration 981 / 9800) loss: 1.561949\n",
            "(Iteration 1001 / 9800) loss: 1.556698\n",
            "(Iteration 1021 / 9800) loss: 1.459553\n",
            "(Iteration 1041 / 9800) loss: 1.338835\n",
            "(Iteration 1061 / 9800) loss: 1.221893\n",
            "(Iteration 1081 / 9800) loss: 1.354312\n",
            "(Iteration 1101 / 9800) loss: 1.545383\n",
            "(Iteration 1121 / 9800) loss: 1.457406\n",
            "(Iteration 1141 / 9800) loss: 1.245830\n",
            "(Iteration 1161 / 9800) loss: 1.541444\n",
            "(Iteration 1181 / 9800) loss: 1.177891\n",
            "(Iteration 1201 / 9800) loss: 1.275794\n",
            "(Iteration 1221 / 9800) loss: 1.486202\n",
            "(Iteration 1241 / 9800) loss: 1.364154\n",
            "(Iteration 1261 / 9800) loss: 1.244790\n",
            "(Iteration 1281 / 9800) loss: 1.345184\n",
            "(Iteration 1301 / 9800) loss: 1.356352\n",
            "(Iteration 1321 / 9800) loss: 1.071911\n",
            "(Iteration 1341 / 9800) loss: 1.410684\n",
            "(Iteration 1361 / 9800) loss: 1.256373\n",
            "(Iteration 1381 / 9800) loss: 1.180638\n",
            "(Iteration 1401 / 9800) loss: 1.493699\n",
            "(Iteration 1421 / 9800) loss: 1.628839\n",
            "(Iteration 1441 / 9800) loss: 1.260059\n",
            "(Iteration 1461 / 9800) loss: 1.617246\n",
            "(Iteration 1481 / 9800) loss: 1.408496\n",
            "(Iteration 1501 / 9800) loss: 1.248702\n",
            "(Iteration 1521 / 9800) loss: 1.307064\n",
            "(Iteration 1541 / 9800) loss: 1.224283\n",
            "(Iteration 1561 / 9800) loss: 1.450108\n",
            "(Iteration 1581 / 9800) loss: 1.275638\n",
            "(Iteration 1601 / 9800) loss: 1.143967\n",
            "(Iteration 1621 / 9800) loss: 1.536701\n",
            "(Iteration 1641 / 9800) loss: 1.099563\n",
            "(Iteration 1661 / 9800) loss: 1.239093\n",
            "(Iteration 1681 / 9800) loss: 1.263083\n",
            "(Iteration 1701 / 9800) loss: 1.522059\n",
            "(Iteration 1721 / 9800) loss: 1.405510\n",
            "(Iteration 1741 / 9800) loss: 1.372411\n",
            "(Iteration 1761 / 9800) loss: 1.079137\n",
            "(Iteration 1781 / 9800) loss: 1.443343\n",
            "(Iteration 1801 / 9800) loss: 1.137319\n",
            "(Iteration 1821 / 9800) loss: 1.322205\n",
            "(Iteration 1841 / 9800) loss: 1.403200\n",
            "(Iteration 1861 / 9800) loss: 1.217773\n",
            "(Iteration 1881 / 9800) loss: 1.381660\n",
            "(Iteration 1901 / 9800) loss: 1.507064\n",
            "(Iteration 1921 / 9800) loss: 1.131200\n",
            "(Iteration 1941 / 9800) loss: 1.558314\n",
            "(Epoch 2 / 10) train acc: 0.665000; val_acc: 0.646000\n",
            "(Iteration 1961 / 9800) loss: 1.257844\n",
            "(Iteration 1981 / 9800) loss: 1.411694\n",
            "(Iteration 2001 / 9800) loss: 1.254242\n",
            "(Iteration 2021 / 9800) loss: 1.635772\n",
            "(Iteration 2041 / 9800) loss: 1.320780\n",
            "(Iteration 2061 / 9800) loss: 1.488415\n",
            "(Iteration 2081 / 9800) loss: 1.116040\n",
            "(Iteration 2101 / 9800) loss: 1.343446\n",
            "(Iteration 2121 / 9800) loss: 1.357931\n",
            "(Iteration 2141 / 9800) loss: 1.293299\n",
            "(Iteration 2161 / 9800) loss: 1.052151\n",
            "(Iteration 2181 / 9800) loss: 1.226448\n",
            "(Iteration 2201 / 9800) loss: 1.469254\n",
            "(Iteration 2221 / 9800) loss: 1.350608\n",
            "(Iteration 2241 / 9800) loss: 1.467261\n",
            "(Iteration 2261 / 9800) loss: 1.528932\n",
            "(Iteration 2281 / 9800) loss: 1.161754\n",
            "(Iteration 2301 / 9800) loss: 1.234186\n",
            "(Iteration 2321 / 9800) loss: 1.370185\n",
            "(Iteration 2341 / 9800) loss: 1.226205\n",
            "(Iteration 2361 / 9800) loss: 1.225161\n",
            "(Iteration 2381 / 9800) loss: 1.292316\n",
            "(Iteration 2401 / 9800) loss: 1.318213\n",
            "(Iteration 2421 / 9800) loss: 1.408417\n",
            "(Iteration 2441 / 9800) loss: 1.253301\n",
            "(Iteration 2461 / 9800) loss: 1.297436\n",
            "(Iteration 2481 / 9800) loss: 1.602112\n",
            "(Iteration 2501 / 9800) loss: 1.708457\n",
            "(Iteration 2521 / 9800) loss: 1.161265\n",
            "(Iteration 2541 / 9800) loss: 1.191401\n",
            "(Iteration 2561 / 9800) loss: 1.385782\n",
            "(Iteration 2581 / 9800) loss: 1.250053\n",
            "(Iteration 2601 / 9800) loss: 1.353212\n",
            "(Iteration 2621 / 9800) loss: 1.352098\n",
            "(Iteration 2641 / 9800) loss: 1.194773\n",
            "(Iteration 2661 / 9800) loss: 1.198116\n",
            "(Iteration 2681 / 9800) loss: 1.065811\n",
            "(Iteration 2701 / 9800) loss: 1.527653\n",
            "(Iteration 2721 / 9800) loss: 1.240495\n",
            "(Iteration 2741 / 9800) loss: 1.383502\n",
            "(Iteration 2761 / 9800) loss: 1.185853\n",
            "(Iteration 2781 / 9800) loss: 1.324703\n",
            "(Iteration 2801 / 9800) loss: 1.186000\n",
            "(Iteration 2821 / 9800) loss: 1.403658\n",
            "(Iteration 2841 / 9800) loss: 1.309621\n",
            "(Iteration 2861 / 9800) loss: 1.114543\n",
            "(Iteration 2881 / 9800) loss: 1.361276\n",
            "(Iteration 2901 / 9800) loss: 1.204573\n",
            "(Iteration 2921 / 9800) loss: 1.253295\n",
            "(Epoch 3 / 10) train acc: 0.706000; val_acc: 0.674000\n",
            "(Iteration 2941 / 9800) loss: 1.352292\n",
            "(Iteration 2961 / 9800) loss: 1.458857\n",
            "(Iteration 2981 / 9800) loss: 1.293781\n",
            "(Iteration 3001 / 9800) loss: 1.171436\n",
            "(Iteration 3021 / 9800) loss: 1.216866\n",
            "(Iteration 3041 / 9800) loss: 0.921628\n",
            "(Iteration 3061 / 9800) loss: 1.460240\n",
            "(Iteration 3081 / 9800) loss: 1.301635\n",
            "(Iteration 3101 / 9800) loss: 1.351044\n",
            "(Iteration 3121 / 9800) loss: 1.217849\n",
            "(Iteration 3141 / 9800) loss: 1.193455\n",
            "(Iteration 3161 / 9800) loss: 1.364903\n",
            "(Iteration 3181 / 9800) loss: 1.299582\n",
            "(Iteration 3201 / 9800) loss: 1.404451\n",
            "(Iteration 3221 / 9800) loss: 1.347897\n",
            "(Iteration 3241 / 9800) loss: 1.294802\n",
            "(Iteration 3261 / 9800) loss: 1.471663\n",
            "(Iteration 3281 / 9800) loss: 1.049167\n",
            "(Iteration 3301 / 9800) loss: 1.393623\n",
            "(Iteration 3321 / 9800) loss: 1.353601\n",
            "(Iteration 3341 / 9800) loss: 1.268088\n",
            "(Iteration 3361 / 9800) loss: 0.988917\n",
            "(Iteration 3381 / 9800) loss: 1.180488\n",
            "(Iteration 3401 / 9800) loss: 1.268512\n",
            "(Iteration 3421 / 9800) loss: 1.372180\n",
            "(Iteration 3441 / 9800) loss: 1.165058\n",
            "(Iteration 3461 / 9800) loss: 1.121798\n",
            "(Iteration 3481 / 9800) loss: 1.264634\n",
            "(Iteration 3501 / 9800) loss: 1.282712\n",
            "(Iteration 3521 / 9800) loss: 1.288701\n",
            "(Iteration 3541 / 9800) loss: 1.230348\n",
            "(Iteration 3561 / 9800) loss: 1.163572\n",
            "(Iteration 3581 / 9800) loss: 1.317622\n",
            "(Iteration 3601 / 9800) loss: 1.308922\n",
            "(Iteration 3621 / 9800) loss: 1.253431\n",
            "(Iteration 3641 / 9800) loss: 1.349402\n",
            "(Iteration 3661 / 9800) loss: 1.163880\n",
            "(Iteration 3681 / 9800) loss: 1.418813\n",
            "(Iteration 3701 / 9800) loss: 1.221531\n",
            "(Iteration 3721 / 9800) loss: 1.348728\n",
            "(Iteration 3741 / 9800) loss: 1.418723\n",
            "(Iteration 3761 / 9800) loss: 1.365916\n",
            "(Iteration 3781 / 9800) loss: 1.197922\n",
            "(Iteration 3801 / 9800) loss: 1.134158\n",
            "(Iteration 3821 / 9800) loss: 1.378779\n",
            "(Iteration 3841 / 9800) loss: 1.551544\n",
            "(Iteration 3861 / 9800) loss: 1.183690\n",
            "(Iteration 3881 / 9800) loss: 1.288023\n",
            "(Iteration 3901 / 9800) loss: 1.337387\n",
            "(Epoch 4 / 10) train acc: 0.742000; val_acc: 0.681000\n",
            "(Iteration 3921 / 9800) loss: 1.196446\n",
            "(Iteration 3941 / 9800) loss: 1.462381\n",
            "(Iteration 3961 / 9800) loss: 1.245104\n",
            "(Iteration 3981 / 9800) loss: 1.556056\n",
            "(Iteration 4001 / 9800) loss: 1.295449\n",
            "(Iteration 4021 / 9800) loss: 1.300713\n",
            "(Iteration 4041 / 9800) loss: 1.117866\n",
            "(Iteration 4061 / 9800) loss: 1.165716\n",
            "(Iteration 4081 / 9800) loss: 1.285129\n",
            "(Iteration 4101 / 9800) loss: 1.172561\n",
            "(Iteration 4121 / 9800) loss: 1.281186\n",
            "(Iteration 4141 / 9800) loss: 1.322654\n",
            "(Iteration 4161 / 9800) loss: 1.326782\n",
            "(Iteration 4181 / 9800) loss: 1.280311\n",
            "(Iteration 4201 / 9800) loss: 1.206207\n",
            "(Iteration 4221 / 9800) loss: 1.266713\n",
            "(Iteration 4241 / 9800) loss: 1.162288\n",
            "(Iteration 4261 / 9800) loss: 1.064231\n",
            "(Iteration 4281 / 9800) loss: 1.380334\n",
            "(Iteration 4301 / 9800) loss: 1.185782\n",
            "(Iteration 4321 / 9800) loss: 1.298891\n",
            "(Iteration 4341 / 9800) loss: 1.195729\n",
            "(Iteration 4361 / 9800) loss: 1.366487\n",
            "(Iteration 4381 / 9800) loss: 1.158233\n",
            "(Iteration 4401 / 9800) loss: 1.110335\n",
            "(Iteration 4421 / 9800) loss: 1.320572\n",
            "(Iteration 4441 / 9800) loss: 1.343708\n",
            "(Iteration 4461 / 9800) loss: 1.347201\n",
            "(Iteration 4481 / 9800) loss: 1.361582\n",
            "(Iteration 4501 / 9800) loss: 1.010107\n",
            "(Iteration 4521 / 9800) loss: 1.039946\n",
            "(Iteration 4541 / 9800) loss: 1.423022\n",
            "(Iteration 4561 / 9800) loss: 1.038822\n",
            "(Iteration 4581 / 9800) loss: 1.438576\n",
            "(Iteration 4601 / 9800) loss: 1.135469\n",
            "(Iteration 4621 / 9800) loss: 1.159833\n",
            "(Iteration 4641 / 9800) loss: 1.190671\n",
            "(Iteration 4661 / 9800) loss: 1.275986\n",
            "(Iteration 4681 / 9800) loss: 1.014229\n",
            "(Iteration 4701 / 9800) loss: 1.460912\n",
            "(Iteration 4721 / 9800) loss: 1.507285\n",
            "(Iteration 4741 / 9800) loss: 1.165539\n",
            "(Iteration 4761 / 9800) loss: 1.047068\n",
            "(Iteration 4781 / 9800) loss: 1.172528\n",
            "(Iteration 4801 / 9800) loss: 1.371903\n",
            "(Iteration 4821 / 9800) loss: 1.318838\n",
            "(Iteration 4841 / 9800) loss: 1.048768\n",
            "(Iteration 4861 / 9800) loss: 1.271946\n",
            "(Iteration 4881 / 9800) loss: 1.380480\n",
            "(Epoch 5 / 10) train acc: 0.761000; val_acc: 0.703000\n",
            "(Iteration 4901 / 9800) loss: 1.440561\n",
            "(Iteration 4921 / 9800) loss: 1.417028\n",
            "(Iteration 4941 / 9800) loss: 1.175329\n",
            "(Iteration 4961 / 9800) loss: 1.208800\n",
            "(Iteration 4981 / 9800) loss: 1.159721\n",
            "(Iteration 5001 / 9800) loss: 1.436386\n",
            "(Iteration 5021 / 9800) loss: 1.372489\n",
            "(Iteration 5041 / 9800) loss: 1.004074\n",
            "(Iteration 5061 / 9800) loss: 1.199239\n",
            "(Iteration 5081 / 9800) loss: 1.307596\n",
            "(Iteration 5101 / 9800) loss: 1.292740\n",
            "(Iteration 5121 / 9800) loss: 1.199128\n",
            "(Iteration 5141 / 9800) loss: 1.079234\n",
            "(Iteration 5161 / 9800) loss: 1.236336\n",
            "(Iteration 5181 / 9800) loss: 1.092950\n",
            "(Iteration 5201 / 9800) loss: 1.127342\n",
            "(Iteration 5221 / 9800) loss: 1.307808\n",
            "(Iteration 5241 / 9800) loss: 1.333915\n",
            "(Iteration 5261 / 9800) loss: 1.125640\n",
            "(Iteration 5281 / 9800) loss: 1.451327\n",
            "(Iteration 5301 / 9800) loss: 1.163725\n",
            "(Iteration 5321 / 9800) loss: 1.196447\n",
            "(Iteration 5341 / 9800) loss: 0.919007\n",
            "(Iteration 5361 / 9800) loss: 1.194097\n",
            "(Iteration 5381 / 9800) loss: 1.390749\n",
            "(Iteration 5401 / 9800) loss: 1.398834\n",
            "(Iteration 5421 / 9800) loss: 1.212644\n",
            "(Iteration 5441 / 9800) loss: 1.106411\n",
            "(Iteration 5461 / 9800) loss: 0.986885\n",
            "(Iteration 5481 / 9800) loss: 1.144331\n",
            "(Iteration 5501 / 9800) loss: 1.271449\n",
            "(Iteration 5521 / 9800) loss: 1.284360\n",
            "(Iteration 5541 / 9800) loss: 1.008185\n",
            "(Iteration 5561 / 9800) loss: 1.300393\n",
            "(Iteration 5581 / 9800) loss: 1.449455\n",
            "(Iteration 5601 / 9800) loss: 1.065495\n",
            "(Iteration 5621 / 9800) loss: 0.934524\n",
            "(Iteration 5641 / 9800) loss: 1.246091\n",
            "(Iteration 5661 / 9800) loss: 0.974333\n",
            "(Iteration 5681 / 9800) loss: 1.248828\n",
            "(Iteration 5701 / 9800) loss: 1.040708\n",
            "(Iteration 5721 / 9800) loss: 1.141503\n",
            "(Iteration 5741 / 9800) loss: 1.402157\n",
            "(Iteration 5761 / 9800) loss: 1.274070\n",
            "(Iteration 5781 / 9800) loss: 1.056090\n",
            "(Iteration 5801 / 9800) loss: 1.246754\n",
            "(Iteration 5821 / 9800) loss: 1.088922\n",
            "(Iteration 5841 / 9800) loss: 1.274805\n",
            "(Iteration 5861 / 9800) loss: 1.090606\n",
            "(Epoch 6 / 10) train acc: 0.742000; val_acc: 0.688000\n",
            "(Iteration 5881 / 9800) loss: 1.568428\n",
            "(Iteration 5901 / 9800) loss: 1.148113\n",
            "(Iteration 5921 / 9800) loss: 1.324237\n",
            "(Iteration 5941 / 9800) loss: 1.306371\n",
            "(Iteration 5961 / 9800) loss: 1.091382\n",
            "(Iteration 5981 / 9800) loss: 1.160343\n",
            "(Iteration 6001 / 9800) loss: 1.371129\n",
            "(Iteration 6021 / 9800) loss: 1.112418\n",
            "(Iteration 6041 / 9800) loss: 0.957434\n",
            "(Iteration 6061 / 9800) loss: 1.193153\n",
            "(Iteration 6081 / 9800) loss: 1.180927\n",
            "(Iteration 6101 / 9800) loss: 1.074544\n",
            "(Iteration 6121 / 9800) loss: 1.115224\n",
            "(Iteration 6141 / 9800) loss: 1.173856\n",
            "(Iteration 6161 / 9800) loss: 1.223527\n",
            "(Iteration 6181 / 9800) loss: 1.289369\n",
            "(Iteration 6201 / 9800) loss: 1.218355\n",
            "(Iteration 6221 / 9800) loss: 1.193655\n",
            "(Iteration 6241 / 9800) loss: 1.013071\n",
            "(Iteration 6261 / 9800) loss: 1.221852\n",
            "(Iteration 6281 / 9800) loss: 1.411608\n",
            "(Iteration 6301 / 9800) loss: 1.479167\n",
            "(Iteration 6321 / 9800) loss: 1.090496\n",
            "(Iteration 6341 / 9800) loss: 1.058328\n",
            "(Iteration 6361 / 9800) loss: 1.288007\n",
            "(Iteration 6381 / 9800) loss: 1.198278\n",
            "(Iteration 6401 / 9800) loss: 1.359172\n",
            "(Iteration 6421 / 9800) loss: 1.384977\n",
            "(Iteration 6441 / 9800) loss: 1.028364\n",
            "(Iteration 6461 / 9800) loss: 1.141787\n",
            "(Iteration 6481 / 9800) loss: 1.191385\n",
            "(Iteration 6501 / 9800) loss: 1.363409\n",
            "(Iteration 6521 / 9800) loss: 1.076800\n",
            "(Iteration 6541 / 9800) loss: 0.997416\n",
            "(Iteration 6561 / 9800) loss: 1.233445\n",
            "(Iteration 6581 / 9800) loss: 1.380330\n",
            "(Iteration 6601 / 9800) loss: 1.050582\n",
            "(Iteration 6621 / 9800) loss: 1.257565\n",
            "(Iteration 6641 / 9800) loss: 1.143548\n",
            "(Iteration 6661 / 9800) loss: 1.226295\n",
            "(Iteration 6681 / 9800) loss: 1.055367\n",
            "(Iteration 6701 / 9800) loss: 1.149838\n",
            "(Iteration 6721 / 9800) loss: 1.260128\n",
            "(Iteration 6741 / 9800) loss: 1.317904\n",
            "(Iteration 6761 / 9800) loss: 1.237491\n",
            "(Iteration 6781 / 9800) loss: 1.243243\n",
            "(Iteration 6801 / 9800) loss: 1.307284\n",
            "(Iteration 6821 / 9800) loss: 1.046402\n",
            "(Iteration 6841 / 9800) loss: 1.202082\n",
            "(Epoch 7 / 10) train acc: 0.771000; val_acc: 0.720000\n",
            "(Iteration 6861 / 9800) loss: 1.275166\n",
            "(Iteration 6881 / 9800) loss: 1.075653\n",
            "(Iteration 6901 / 9800) loss: 1.120474\n",
            "(Iteration 6921 / 9800) loss: 1.394484\n",
            "(Iteration 6941 / 9800) loss: 1.170814\n",
            "(Iteration 6961 / 9800) loss: 1.185378\n",
            "(Iteration 6981 / 9800) loss: 1.451868\n",
            "(Iteration 7001 / 9800) loss: 1.068697\n",
            "(Iteration 7021 / 9800) loss: 1.075122\n",
            "(Iteration 7041 / 9800) loss: 1.194754\n",
            "(Iteration 7061 / 9800) loss: 1.396497\n",
            "(Iteration 7081 / 9800) loss: 1.213966\n",
            "(Iteration 7101 / 9800) loss: 1.332992\n",
            "(Iteration 7121 / 9800) loss: 1.283425\n",
            "(Iteration 7141 / 9800) loss: 1.178670\n",
            "(Iteration 7161 / 9800) loss: 1.169049\n",
            "(Iteration 7181 / 9800) loss: 1.313241\n",
            "(Iteration 7201 / 9800) loss: 0.995257\n",
            "(Iteration 7221 / 9800) loss: 1.246492\n",
            "(Iteration 7241 / 9800) loss: 1.058619\n",
            "(Iteration 7261 / 9800) loss: 1.414958\n",
            "(Iteration 7281 / 9800) loss: 1.275821\n",
            "(Iteration 7301 / 9800) loss: 1.152308\n",
            "(Iteration 7321 / 9800) loss: 1.214756\n",
            "(Iteration 7341 / 9800) loss: 1.328622\n",
            "(Iteration 7361 / 9800) loss: 1.270010\n",
            "(Iteration 7381 / 9800) loss: 1.486572\n",
            "(Iteration 7401 / 9800) loss: 1.280191\n",
            "(Iteration 7421 / 9800) loss: 1.549231\n",
            "(Iteration 7441 / 9800) loss: 1.348715\n",
            "(Iteration 7461 / 9800) loss: 1.324584\n",
            "(Iteration 7481 / 9800) loss: 0.962346\n",
            "(Iteration 7501 / 9800) loss: 1.049797\n",
            "(Iteration 7521 / 9800) loss: 1.360466\n",
            "(Iteration 7541 / 9800) loss: 1.126431\n",
            "(Iteration 7561 / 9800) loss: 1.020592\n",
            "(Iteration 7581 / 9800) loss: 1.209832\n",
            "(Iteration 7601 / 9800) loss: 1.165428\n",
            "(Iteration 7621 / 9800) loss: 1.407695\n",
            "(Iteration 7641 / 9800) loss: 1.181394\n",
            "(Iteration 7661 / 9800) loss: 1.271089\n",
            "(Iteration 7681 / 9800) loss: 1.110779\n",
            "(Iteration 7701 / 9800) loss: 1.215682\n",
            "(Iteration 7721 / 9800) loss: 1.009951\n",
            "(Iteration 7741 / 9800) loss: 1.289112\n",
            "(Iteration 7761 / 9800) loss: 0.987200\n",
            "(Iteration 7781 / 9800) loss: 1.281331\n",
            "(Iteration 7801 / 9800) loss: 1.198360\n",
            "(Iteration 7821 / 9800) loss: 1.173311\n",
            "(Epoch 8 / 10) train acc: 0.799000; val_acc: 0.728000\n",
            "(Iteration 7841 / 9800) loss: 1.070520\n",
            "(Iteration 7861 / 9800) loss: 1.112105\n",
            "(Iteration 7881 / 9800) loss: 1.502476\n",
            "(Iteration 7901 / 9800) loss: 1.121273\n",
            "(Iteration 7921 / 9800) loss: 1.207193\n",
            "(Iteration 7941 / 9800) loss: 1.081277\n",
            "(Iteration 7961 / 9800) loss: 1.072658\n",
            "(Iteration 7981 / 9800) loss: 1.117615\n",
            "(Iteration 8001 / 9800) loss: 1.251970\n",
            "(Iteration 8021 / 9800) loss: 1.011409\n",
            "(Iteration 8041 / 9800) loss: 1.184763\n",
            "(Iteration 8061 / 9800) loss: 1.148202\n",
            "(Iteration 8081 / 9800) loss: 1.184252\n",
            "(Iteration 8101 / 9800) loss: 1.114231\n",
            "(Iteration 8121 / 9800) loss: 1.276033\n",
            "(Iteration 8141 / 9800) loss: 1.267164\n",
            "(Iteration 8161 / 9800) loss: 1.064479\n",
            "(Iteration 8181 / 9800) loss: 1.330041\n",
            "(Iteration 8201 / 9800) loss: 1.298767\n",
            "(Iteration 8221 / 9800) loss: 1.128041\n",
            "(Iteration 8241 / 9800) loss: 1.070679\n",
            "(Iteration 8261 / 9800) loss: 1.293693\n",
            "(Iteration 8281 / 9800) loss: 1.286268\n",
            "(Iteration 8301 / 9800) loss: 0.993477\n",
            "(Iteration 8321 / 9800) loss: 1.218805\n",
            "(Iteration 8341 / 9800) loss: 1.194090\n",
            "(Iteration 8361 / 9800) loss: 0.991985\n",
            "(Iteration 8381 / 9800) loss: 1.015665\n",
            "(Iteration 8401 / 9800) loss: 1.490396\n",
            "(Iteration 8421 / 9800) loss: 1.192844\n",
            "(Iteration 8441 / 9800) loss: 1.142493\n",
            "(Iteration 8461 / 9800) loss: 1.025923\n",
            "(Iteration 8481 / 9800) loss: 1.293122\n",
            "(Iteration 8501 / 9800) loss: 1.034668\n",
            "(Iteration 8521 / 9800) loss: 1.144350\n",
            "(Iteration 8541 / 9800) loss: 0.919091\n",
            "(Iteration 8561 / 9800) loss: 1.052981\n",
            "(Iteration 8581 / 9800) loss: 1.305136\n",
            "(Iteration 8601 / 9800) loss: 1.001497\n",
            "(Iteration 8621 / 9800) loss: 1.401286\n",
            "(Iteration 8641 / 9800) loss: 1.152726\n",
            "(Iteration 8661 / 9800) loss: 1.199686\n",
            "(Iteration 8681 / 9800) loss: 0.972454\n",
            "(Iteration 8701 / 9800) loss: 1.373823\n",
            "(Iteration 8721 / 9800) loss: 1.082185\n",
            "(Iteration 8741 / 9800) loss: 1.213235\n",
            "(Iteration 8761 / 9800) loss: 1.515653\n",
            "(Iteration 8781 / 9800) loss: 1.290446\n",
            "(Iteration 8801 / 9800) loss: 0.828159\n",
            "(Epoch 9 / 10) train acc: 0.801000; val_acc: 0.723000\n",
            "(Iteration 8821 / 9800) loss: 1.248472\n",
            "(Iteration 8841 / 9800) loss: 0.930644\n",
            "(Iteration 8861 / 9800) loss: 1.335992\n",
            "(Iteration 8881 / 9800) loss: 1.087816\n",
            "(Iteration 8901 / 9800) loss: 1.213428\n",
            "(Iteration 8921 / 9800) loss: 0.977476\n",
            "(Iteration 8941 / 9800) loss: 1.228507\n",
            "(Iteration 8961 / 9800) loss: 1.187446\n",
            "(Iteration 8981 / 9800) loss: 1.088277\n",
            "(Iteration 9001 / 9800) loss: 1.006421\n",
            "(Iteration 9021 / 9800) loss: 1.221866\n",
            "(Iteration 9041 / 9800) loss: 1.154176\n",
            "(Iteration 9061 / 9800) loss: 0.929970\n",
            "(Iteration 9081 / 9800) loss: 1.191865\n",
            "(Iteration 9101 / 9800) loss: 1.112166\n",
            "(Iteration 9121 / 9800) loss: 1.031140\n",
            "(Iteration 9141 / 9800) loss: 1.236948\n",
            "(Iteration 9161 / 9800) loss: 1.116866\n",
            "(Iteration 9181 / 9800) loss: 1.044033\n",
            "(Iteration 9201 / 9800) loss: 1.154068\n",
            "(Iteration 9221 / 9800) loss: 1.525877\n",
            "(Iteration 9241 / 9800) loss: 1.128745\n",
            "(Iteration 9261 / 9800) loss: 1.564470\n",
            "(Iteration 9281 / 9800) loss: 1.264303\n",
            "(Iteration 9301 / 9800) loss: 1.088406\n",
            "(Iteration 9321 / 9800) loss: 1.001677\n",
            "(Iteration 9341 / 9800) loss: 0.994601\n",
            "(Iteration 9361 / 9800) loss: 1.211796\n",
            "(Iteration 9381 / 9800) loss: 1.166386\n",
            "(Iteration 9401 / 9800) loss: 0.996197\n",
            "(Iteration 9421 / 9800) loss: 1.029850\n",
            "(Iteration 9441 / 9800) loss: 1.067479\n",
            "(Iteration 9461 / 9800) loss: 1.033508\n",
            "(Iteration 9481 / 9800) loss: 1.257158\n",
            "(Iteration 9501 / 9800) loss: 1.368695\n",
            "(Iteration 9521 / 9800) loss: 1.308272\n",
            "(Iteration 9541 / 9800) loss: 1.205890\n",
            "(Iteration 9561 / 9800) loss: 1.094590\n",
            "(Iteration 9581 / 9800) loss: 1.307483\n",
            "(Iteration 9601 / 9800) loss: 1.160815\n",
            "(Iteration 9621 / 9800) loss: 1.132547\n",
            "(Iteration 9641 / 9800) loss: 1.247550\n",
            "(Iteration 9661 / 9800) loss: 1.279765\n",
            "(Iteration 9681 / 9800) loss: 1.153025\n",
            "(Iteration 9701 / 9800) loss: 1.283367\n",
            "(Iteration 9721 / 9800) loss: 1.324911\n",
            "(Iteration 9741 / 9800) loss: 1.092842\n",
            "(Iteration 9761 / 9800) loss: 1.241338\n",
            "(Iteration 9781 / 9800) loss: 1.364096\n",
            "(Epoch 10 / 10) train acc: 0.820000; val_acc: 0.736000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "VhS08_zIvnTR",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# model = MultiLayerConvNet(weight_scale=0.001, hidden_dim=[800,400,200], reg=0.001, dropout = 0.5)\n",
        "\n",
        "# solver = Solver(model, data,\n",
        "#                 num_epochs=10, batch_size=50,\n",
        "#                 update_rule='adam',\n",
        "#                 optim_config={\n",
        "#                   'learning_rate': 1e-3,\n",
        "#                 },\n",
        "#                 verbose=True, print_every=20)\n",
        "# solver.train()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ldWbhItsPJwi",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "vc9YbQsUQad-",
        "colab_type": "code",
        "outputId": "695eebec-035e-424d-c9b7-0bcc0f8abd9a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 499
        }
      },
      "cell_type": "code",
      "source": [
        "plt.subplot(2, 1, 1)\n",
        "plt.plot(solver.loss_history, 'o')\n",
        "plt.xlabel('iteration')\n",
        "plt.ylabel('loss')\n",
        "\n",
        "plt.subplot(2, 1, 2)\n",
        "plt.plot(solver.train_acc_history, '-o')\n",
        "plt.plot(solver.val_acc_history, '-o')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.xlabel('epoch')\n",
        "plt.ylabel('accuracy')\n",
        "plt.show()"
      ],
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAHiCAYAAADWA6krAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 720x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "1lEQeBCUQvEa",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "okMDZAcoQu4L",
        "colab_type": "code",
        "outputId": "1852f100-f92f-4625-fbd9-3709627e65c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        }
      },
      "cell_type": "code",
      "source": [
        "from cs231n.vis_utils import visualize_grid\n",
        "\n",
        "grid = visualize_grid(model.params['W1'].transpose(0, 2, 3, 1))\n",
        "plt.imshow(grid.astype('uint8'))\n",
        "plt.axis('off')\n",
        "plt.gcf().set_size_inches(5, 5)\n",
        "plt.show()"
      ],
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATAAAAEvCAYAAADCe529AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGFCWgAFlCRhQloABZQkYUNZ/A/DDBiNYGb3rAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "f58nVi7Adx2v",
        "colab_type": "code",
        "outputId": "5b2c00d1-745b-41c7-f800-c48ef657d678",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "solver.check_accuracy(data['X_test'], data['y_test'])"
      ],
      "execution_count": 122,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.727"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 122
        }
      ]
    },
    {
      "metadata": {
        "id": "Qj_AirDWd-pV",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "tcqK6Cxnfuna",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "wZRhBDxqfu8u",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "def dropout_forward(x, dropout_param):\n",
        "  \"\"\"\n",
        "  Performs the forward pass for (inverted) dropout.\n",
        "\n",
        "  Inputs:\n",
        "  - x: Input data, of any shape\n",
        "  - dropout_param: A dictionary with the following keys:\n",
        "    - p: Dropout parameter. We drop each neuron output with probability p.\n",
        "    - mode: 'test' or 'train'. If the mode is train, then perform dropout;\n",
        "      if the mode is test, then just return the input.\n",
        "    - seed: Seed for the random number generator. Passing seed makes this\n",
        "      function deterministic, which is needed for gradient checking but not in\n",
        "      real networks.\n",
        "\n",
        "  Outputs:\n",
        "  - out: Array of the same shape as x.\n",
        "  - cache: A tuple (dropout_param, mask). In training mode, mask is the dropout\n",
        "    mask that was used to multiply the input; in test mode, mask is None.\n",
        "  \"\"\"\n",
        "  p, mode = dropout_param['p'], dropout_param['mode']\n",
        "  if 'seed' in dropout_param:\n",
        "    np.random.seed(dropout_param['seed'])\n",
        "\n",
        "  mask = None\n",
        "  out = None\n",
        "\n",
        "  if mode == 'train':\n",
        "    ###########################################################################\n",
        "    # TODO: Implement the training phase forward pass for inverted dropout.   #\n",
        "    # Store the dropout mask in the mask variable.                            #\n",
        "    ###########################################################################\n",
        "    #pass\n",
        "    #mask = (np.random.rand((x.shape[0],x.shape[1]))<p)/p\n",
        "    #print type(x.shape)\n",
        "    \n",
        "    mask = (np.random.random_sample(x.shape)<p)/p\n",
        "    out = x\n",
        "    ###########################################################################\n",
        "    #                            END OF YOUR CODE                             #\n",
        "    ###########################################################################\n",
        "  elif mode == 'test':\n",
        "    ###########################################################################\n",
        "    # TODO: Implement the test phase forward pass for inverted dropout.       #\n",
        "    ###########################################################################\n",
        "    #pass\n",
        "    mask = np.ones((x.shape))\n",
        "    out = x\n",
        "    ###########################################################################\n",
        "    #                            END OF YOUR CODE                             #\n",
        "    ###########################################################################\n",
        "\n",
        "  cache = (dropout_param, mask)\n",
        "  out = out.astype(x.dtype, copy=False)\n",
        "\n",
        "  return out, cache\n",
        "\n",
        "\n",
        "def dropout_backward(dout, cache):\n",
        "  \"\"\"\n",
        "  Perform the backward pass for (inverted) dropout.\n",
        "\n",
        "  Inputs:\n",
        "  - dout: Upstream derivatives, of any shape\n",
        "  - cache: (dropout_param, mask) from dropout_forward.\n",
        "  \"\"\"\n",
        "  dropout_param, mask = cache\n",
        "  mode = dropout_param['mode']\n",
        "  \n",
        "  dx = None\n",
        "  if mode == 'train':\n",
        "    ###########################################################################\n",
        "    # TODO: Implement the training phase backward pass for inverted dropout.  #\n",
        "    ###########################################################################\n",
        "    #pass\n",
        "    p = dropout_param['p']\n",
        "    dx = dout*mask\n",
        "    ###########################################################################\n",
        "    #                            END OF YOUR CODE                             #\n",
        "    ###########################################################################\n",
        "  elif mode == 'test':\n",
        "    dx = dout\n",
        "  return dx\n"
      ]
    },
    {
      "metadata": {
        "id": "VRMC_e69fyGM",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# def dropout_forward(x, dropout_param):\n",
        "#   \"\"\"\n",
        "#   Performs the forward pass for (inverted) dropout.\n",
        "\n",
        "#   Inputs:\n",
        "#   - x: Input data, of any shape\n",
        "#   - dropout_param: A dictionary with the following keys:\n",
        "#     - p: Dropout parameter. We drop each neuron output with probability p.\n",
        "#     - mode: 'test' or 'train'. If the mode is train, then perform dropout;\n",
        "#       if the mode is test, then just return the input.\n",
        "#     - seed: Seed for the random number generator. Passing seed makes this\n",
        "#       function deterministic, which is needed for gradient checking but not in\n",
        "#       real networks.\n",
        "\n",
        "#   Outputs:\n",
        "#   - out: Array of the same shape as x.\n",
        "#   - cache: A tuple (dropout_param, mask). In training mode, mask is the dropout\n",
        "#     mask that was used to multiply the input; in test mode, mask is None.\n",
        "#   \"\"\"\n",
        "#   p, mode = dropout_param['p'], dropout_param['mode']\n",
        "#   if 'seed' in dropout_param:\n",
        "#     np.random.seed(dropout_param['seed'])\n",
        "\n",
        "#   mask = None\n",
        "#   out = None\n",
        "\n",
        "#   if mode == 'train':\n",
        "#     ###########################################################################\n",
        "#     # TODO: Implement the training phase forward pass for inverted dropout.   #\n",
        "#     # Store the dropout mask in the mask variable.                            #\n",
        "#     ###########################################################################\n",
        "#     #pass\n",
        "#     #mask = (np.random.rand((x.shape[0],x.shape[1]))<p)/p\n",
        "#     #print type(x.shape)\n",
        "    \n",
        "#     mask = (np.random.random_sample(x.shape)<p)/p\n",
        "#     out = x\n",
        "#     ###########################################################################\n",
        "#     #                            END OF YOUR CODE                             #\n",
        "#     ###########################################################################\n",
        "#   elif mode == 'test':\n",
        "#     ###########################################################################\n",
        "#     # TODO: Implement the test phase forward pass for inverted dropout.       #\n",
        "#     ###########################################################################\n",
        "#     #pass\n",
        "#     mask = np.ones((x.shape))\n",
        "#     out = x\n",
        "#     ###########################################################################\n",
        "#     #                            END OF YOUR CODE                             #\n",
        "#     ###########################################################################\n",
        "\n",
        "#   cache = (dropout_param, mask)\n",
        "#   out = out.astype(x.dtype, copy=False)\n",
        "\n",
        "#   return out, cache\n",
        "\n",
        "\n",
        "# def dropout_backward(dout, cache):\n",
        "#   \"\"\"\n",
        "#   Perform the backward pass for (inverted) dropout.\n",
        "\n",
        "#   Inputs:\n",
        "#   - dout: Upstream derivatives, of any shape\n",
        "#   - cache: (dropout_param, mask) from dropout_forward.\n",
        "#   \"\"\"\n",
        "#   dropout_param, mask = cache\n",
        "#   mode = dropout_param['mode']\n",
        "  \n",
        "#   dx = None\n",
        "#   if mode == 'train':\n",
        "#     ###########################################################################\n",
        "#     # TODO: Implement the training phase backward pass for inverted dropout.  #\n",
        "#     ###########################################################################\n",
        "#     #pass\n",
        "#     p = dropout_param['p']\n",
        "#     dx = dout*mask\n",
        "#     ###########################################################################\n",
        "#     #                            END OF YOUR CODE                             #\n",
        "#     ###########################################################################\n",
        "#   elif mode == 'test':\n",
        "#     dx = dout\n",
        "#   return dx\n"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}