lesson7_resnet_mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "lesson7-resnet-mnist.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A3s4IYLA5zxS",
        "colab_type": "text"
      },
      "source": [
        "## MNIST CNN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jc3FeCby5zxV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%reload_ext autoreload\n",
        "%autoreload 2\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jhk15XbT5zxa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from fastai.vision import *"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "84Ly4rVg5zxd",
        "colab_type": "text"
      },
      "source": [
        "### Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K5wqdTwH5zxd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "path = untar_data(URLs.MNIST)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KideuA5m5zxg",
        "colab_type": "code",
        "outputId": "827e2a5d-d97d-4939-9bdc-41831e16f52a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "path.ls()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[PosixPath('/root/.fastai/data/mnist_png/testing'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training')]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l1tQH5o65zxm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "il = ImageList.from_folder(path, convert_mode='L')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uW72QW_H5zxp",
        "colab_type": "code",
        "outputId": "c3f85d53-5ad8-467e-f234-6fba3126fb23",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "il.items[0]"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "PosixPath('/root/.fastai/data/mnist_png/testing/8/6850.png')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "atIHX0nk5zxt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "defaults.cmap='binary'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EkxBVgbd5zxx",
        "colab_type": "code",
        "outputId": "fef92612-c65d-4093-fb42-797d9bbfd80c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "il"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ImageList (70000 items)\n",
              "Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28)\n",
              "Path: /root/.fastai/data/mnist_png"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zF4dIlo85zx1",
        "colab_type": "code",
        "outputId": "82310106-2e88-42d4-dbe9-da7114c66f44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 212
        }
      },
      "source": [
        "il[0].show()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAADDCAYAAAAyYdXtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 216x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y4vkK0pO5zx6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "sd = il.split_by_folder(train='training', valid='testing')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "US8QpyZu5zx9",
        "colab_type": "code",
        "outputId": "053bce10-8596-47ab-f213-5f91e35dbc9b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "sd"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ItemLists;\n",
              "\n",
              "Train: ImageList (60000 items)\n",
              "Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28)\n",
              "Path: /root/.fastai/data/mnist_png;\n",
              "\n",
              "Valid: ImageList (10000 items)\n",
              "Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28)\n",
              "Path: /root/.fastai/data/mnist_png;\n",
              "\n",
              "Test: None"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GO3r-Dwu5zyB",
        "colab_type": "code",
        "outputId": "3383a97e-8a6c-4e16-8f51-4ea4a28f5ab8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "source": [
        "(path/'training').ls()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[PosixPath('/root/.fastai/data/mnist_png/training/8'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/9'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/1'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/0'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/6'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/4'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/7'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/2'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/5'),\n",
              " PosixPath('/root/.fastai/data/mnist_png/training/3')]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "szhO-Gv85zyF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ll = sd.label_from_folder()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oLb29fvE5zyK",
        "colab_type": "code",
        "outputId": "18b19861-464c-4bb2-cd30-6203948d81b1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        }
      },
      "source": [
        "ll"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LabelLists;\n",
              "\n",
              "Train: LabelList (60000 items)\n",
              "x: ImageList\n",
              "Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28)\n",
              "y: CategoryList\n",
              "8,8,8,8,8\n",
              "Path: /root/.fastai/data/mnist_png;\n",
              "\n",
              "Valid: LabelList (10000 items)\n",
              "x: ImageList\n",
              "Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28),Image (1, 28, 28)\n",
              "y: CategoryList\n",
              "8,8,8,8,8\n",
              "Path: /root/.fastai/data/mnist_png;\n",
              "\n",
              "Test: None"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JGV-RZVg5zyO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x,y = ll.train[0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZDa3jm7D5zyS",
        "colab_type": "code",
        "outputId": "b38cd9bb-4e61-4644-ab2f-f8bb224475dc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 229
        }
      },
      "source": [
        "x.show()\n",
        "print(y,x.shape)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "8 torch.Size([1, 28, 28])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAADDCAYAAAAyYdXtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 216x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9ypuEEq45zyY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tfms = ([*rand_pad(padding=3, size=28, mode='zeros')], [])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dmKykJUV5zyc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ll = ll.transform(tfms)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j-rxyiJh5zyh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "bs = 128"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ah83vrYn5zyk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# not using imagenet_stats because not using pretrained model\n",
        "data = ll.databunch(bs=bs).normalize()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HADdICUQ5zyt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x,y = data.train_ds[0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y2GXjBse5zyv",
        "colab_type": "code",
        "outputId": "d29f7cea-7308-4a3e-a6ee-97cc91d08eba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 229
        }
      },
      "source": [
        "x.show()\n",
        "print(y)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "8\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMUAAADDCAYAAAAyYdXtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 216x216 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "skIshZpA5zy0",
        "colab_type": "code",
        "outputId": "6efcb91f-6b7e-41b1-af1a-a5bb6f248dce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 484
        }
      },
      "source": [
        "def _plot(i,j,ax): data.train_ds[0][0].show(ax, cmap='gray')\n",
        "plot_multi(_plot, 3, 3, figsize=(8,8))"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd0AAAHTCAYAAABiN8IeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAGNVJREFUeJzt3c2rzd3/P/C1L3dxhESKGDm5SZkY\nmBImlwkTRCaUGEkpwj9BKTMloXRlYCbKRMpEjNSRko7c5i6ndNjfwdXV7/de6/1x9tn2fp199n48\nZmu1zt67vK/9vFbrtV670Ww2EwDQfX9N9QcAgEEhdAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIDMj\n36zRaLgUPMCazWZjKt7XczfYpuK588wNtt89c3a6ABBE6AJAEKELAEGELgAEEboAEEToAkAQoQsA\nQYQuAAQJbY7B1Prw4UMx9+nTp8p43bp1xZofP3507TMBDBI7XQAIInQBIIjQBYAgQhcAgiikCjRv\n3rxibvfu3b8dp5TS8PBwR95/4cKFxdzixYsr45kzy0dCIRXQK/KC0LwYNKXeLgi10wWAIEIXAIII\nXQAI4ky3QzZt2lQZnzhxolizd+/eYq7RaHTtM7Xi5s2blXGvnHswWDRumT7y2pS6OpTI2pS8LiWl\n3q5NsdMFgCBCFwCCCF0ACCJ0ASCIQqoWDA0NVcYnT54s1hw/frwyrmtE0a68oGR8fLxYkxcT/PVX\na/8/dfHixQlfm/6ncQsptVcQ2mvFoCn19nNhpwsAQYQuAAQRugAQxJluC/Jzz4MHD7b1Ol++fCnm\n8jOTuiYBd+/erYy/fv1arHn58mVlvHLlymLNo0ePirmnT5/Wf1j6Rn5Ol1L53Gnc0v/6tTYl/37+\nX6/dK+x0ASCI0AWAIEIXAIIIXQAIopAqs2bNmmKurilALj+4v3//frFm3759xdy7d+8m8en+tXPn\nzmJuxYoVE/7dmTNnirn3799P+v3pHe0Ux6TUuQKZdopjUmqvQKaXi2Omg24VhNY10IgsCJ1uxaB2\nugAQROgCQBChCwBBhC4ABFFIlan71ZT58+dP+HcjIyOV8bZt2zr2mWbPnl0Z1xVE5YUpo6OjxZqH\nDx927DPRG6Zjt7SUygIZ3dI6K7IgtJ1i0JQ6VxA63YpB7XQBIIjQBYAgQhcAgjjT7ZDz58937bVP\nnTpVGW/evLlYk5+lbd26dcI1TC+dOqdLqTyr07ilv6hN6V12ugAQROgCQBChCwBBhC4ABFFIlXn1\n6lUxNzY2VhnPnTu3WNOpIqWzZ88Wc3khVZ1r165Vxs+ePevI56F3dKo4JqXOFci0UxyTUlkgM92L\nY/qFgtDus9MFgCBCFwCCCF0ACOJMN/P48eNi7vnz55Xxhg0bijXLli1r6/2uXLlSGe/fv79Y02g0\nKuPr168Xa44dO9bW+9P/eu2cLqXyrG66n9P1GrUpvctOFwCCCF0ACCJ0ASCI0AWAIAqpWnDr1q3K\nuK6QKm8AsH79+mJNXeOAXbt2Tfj+p0+frowvXbpUrPn58+eEr8P01i/FMSn1Z4FML4ksCM2LQVNS\nEPo7droAEEToAkAQoQsAQZzpdkh+3nXy5MlizdKlS4u5ZrNZGR8/frxY083mBkwfGrfwJ7pVm9JK\nXUpKalP+Y6cLAEGELgAEEboAEEToAkAQhVSZoaGhYm7jxo0T/t2CBQsmXFPX3GDPnj2V8YMHDyZ8\nHfhPO8UxKZUFMhq3kFJ7BaF5MWhKCkJ/x04XAIIIXQAIInQBIIjQBYAgA19ItX379sr48uXLxZrl\ny5dP+nVHR0eLuS1bthRzIyMjk35tmIy6XwfKC2R0S+svkQWheTFoSgpCf8dOFwCCCF0ACCJ0ASBI\n357pzpo1q5i7cOFCMXf48OHKeMaMGcWab9++Vcbj4+PFmkWLFlXGd+7cKdY4v+VPaNxCnbwuJaXY\n2hTfa5NjpwsAQYQuAAQRugAQROgCQJC+LaQaHh4u5o4cOTLh371586aY27FjR2W8du3aYs2NGzcq\n41WrVhVr6n7t5devXxN+JgZTtxq3pFQWyGjc0ptaKQjNi0FTUhDay+x0ASCI0AWAIEIXAIL07Zlu\n3YXxOvmZat35yJMnTyrjujPdXN0ZWd05izPdwZSf1UU2bkmpPKtzTtebpkNtiu+wybHTBYAgQhcA\ngghdAAgidAEgSN8WUrVaGPLx48fK+Pbt28Wa/NddDh06NOHrvnjxophTcMB/8gKZyOKYlMoCGY1b\nelMrBaF1/06RBaGek8mx0wWAIEIXAIIIXQAI0rdnuvfu3SvmPn/+XMwtWbKkMv7+/XuxJj/vmjNn\nTrHm7du3lXHdGd3Pnz/rPywDp52zuk6d06VUntVp3NKbWqlNyetSUlKb0svsdAEgiNAFgCBCFwCC\nCF0ACNK3hVRjY2PF3KZNm4q5EydOVMZHjx6d8LXrigvOnTtXGee/4gL/v3YKZDpVHJNS+QwrjulN\nrRSE5sWgKSkI7WV2ugAQROgCQBChCwBBhC4ABGk0m824N2s04t6MntNsNhtT8b69+NzNnTu3Mn79\n+nWxZuHChZVxXXFgO8UxKaV04MCByrifC/+m4rnr5jO3evXqyjgvBk2pcwWhV69eneSnI6XfP3N2\nugAQROgCQBChCwBBnOkSxpnu/5af06XUvcYtKQ3WWV2/nenS+5zpAkAPELoAEEToAkAQoQsAQRRS\nEUYhFVNBIRXRFFIBQA8QugAQROgCQBChCwBBhC4ABBG6ABBE6AJAEKELAEGELgAEEboAEEToAkAQ\noQsAQYQuAAQRugAQROgCQBChCwBBhC4ABGk0m82p/gwAMBDsdAEgiNAFgCBCFwCCCF0ACCJ0ASCI\n0AWAIEIXAIIIXQAIInQBIIjQBYAgQhcAgghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIX\nAIIIXQAIInQBIIjQBYAgQhcAgghdAAgyM/LNGo1GM/L96C3NZrMxFe/ruRtsU/HceeYG2++eOTtd\nAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIXAIIIXQAIInQBIIjQBYAgQhcAgghdAAgidAEg\niNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIXAIIIXQAIInQBIMjMqf4AANALPnz4UBl/+vSpWLNu3brK\n+MePH5N6DztdAAgidAEgiNAFgCDOdAHoCfPmzauMd+/eXazJ54aHhzv2/gsXLqyMFy9eXKyZObMa\nm850AaBHCV0ACCJ0ASCI0AWAIAqpBkh+8Tul8vJ3fvE7pckXCgDkNm3aVBmfOHGiWLN3797KuNFo\ndPUzTeTmzZvF3J9+H9rpAkAQoQsAQYQuAAQRugAQRCFVoLzbSkpld5W6Diyd6riSd1tJqey4kndb\nSUkhFZ0X8WsuxBgaGirmTp48WcwdP368Mq77PmpH3bMzPj5ezOXfdX/9NfGe8+LFiy299mTY6QJA\nEKELAEGELgAEcabbIe1c/E6p9y5/OzcbDIPway7EqDv3PHjwYFuv9eXLl8q47ns0rwe4e/dusebr\n16/F3MuXLyvjlStXFmsePXpUGT99+vR/f9g22ekCQBChCwBBhC4ABBG6ABBEIVUL8svfkRe/Uyov\nf3fq4ndKZRHEn178ZmrlBX0pDe6vudAda9asqYzrivDq5N8t9+/fL9bs27evMn737t0kP92/du7c\nWcytWLFiwr87c+ZMZfz+/fu23v937HQBIIjQBYAgQhcAgjjTbUF+7tmpi98pledt+cXvlMrL3526\n+J1Sdy5/0x2tNJbPawtSim0sX9fkYqoay9MdeZOU+fPnt/R3IyMjlfG2bds68nlmz55dzOVnsymV\nz+Ho6Gix5uHDhx35TL9jpwsAQYQuAAQRugAQROgCQBCFVJn84ndKrV3+bufid0rtXf7u1MXvlLpz\n+ZvumA6/5pIX9KU0db/mQm85f/58V1731KlTxdzmzZuLufxZ3bp164RrusFOFwCCCF0ACCJ0ASCI\nM91MfvE7pdYuf3fr4ndK5eXvXr74Tee001i+rqFEZGP5VmoLUoppLE93vHr1qjIeGxsr1sydO7eY\n69R56dmzZyvjujPdOteuXauMnz171pHPM1l2ugAQROgCQBChCwBBhC4ABFFI1SHduvidUlko0MsX\nv+mcdn7NJS/oSyn211zqflFIUV9/efz4cWX8/PnzYs2GDRuKuWXLlk36va5cuVLM7d+/vzJuNBrF\nmuvXrxdzx44dm/T7d4OdLgAEEboAEEToAkAQoQsAQRRSZfJuKymVHVciu62k1FrHlV7ptsLUiizo\nS6ks6qv770BRX3+7detWMVdXSJUX2a1fv75Ykxfm7dq1a8L3P336dDF36dKlYu7nz58TvlYEO10A\nCCJ0ASCI0AWAIM50M/nF75TKy9+duvidUnn5O7/4nVJ5+buXL37TOe38mksnz0rb+TWXvLYgJfUF\n/Ct/nk6ePFmsWbp0aWXcbDaLNcePH6+Mu1nH0A12ugAQROgCQBChCwBBhC4ABFFI1YL88nenLn6n\n1N7l716++E3ntPNrLp0q6EupvV9zUdDX/4aGhirjjRs3tvR3CxYsmHBNXjy4Z8+eYs2DBw9aer9e\nZacLAEGELgAEEboAEMSZboe0c/E7pfLyd37xO6Xpd/mb7milsXxeW5BSbGN5tQX9Zfv27cXc5cuX\nK+Ply5e39dqjo6PF3JYtWyrjkZGRtl67l9npAkAQoQsAQYQuAAQRugAQRCFVJr/4nVJrl7/bufid\nUnn5e7pf/GZq5QV9KQ3ur7nwe7NmzSrmLly4UBkfPny4WDNjxozK+Nu3b8Wa8fHxYm7RokWV8Z07\nd4o1/Vg4lbPTBYAgQhcAgghdAAgy8Ge6+eXv/OJ3Su1d/m7l4ndKg3GGQXvaaSzfSm1BSoPRWJ7f\nGx4eLuaOHDky4d+9efOmMt6xY0exZu3atcXcjRs3KuNVq1YVa/LmLr9+/Zrw80w3droAEEToAkAQ\noQsAQYQuAATp20KqVi5+p1Re/s4vfqdUXv528ZtO82suRKt75nJ1hUz5d+aTJ0+KNXWFVLm6wtL8\n+1chFQDQNqELAEGELgAEEboAEKRvC6k61W0lpbLjSqe6raTUn4UCVEX+mkte0JeSoj7qtfIMfPz4\nsZi7fft2ZVz3y2yHDh2a8LVfvHhRzA3C96GdLgAEEboAEEToAkCQvj3TbeXid0rlGULd2Vp++btT\nF7/r3p/+E/lrLnltQUqD+2su/N69e/eKuc+fP1fGS5YsKdZ8//69Mq6rVZkzZ04x9/bt28q47r+B\nnz9/1n/YPmKnCwBBhC4ABBG6ABBE6AJAkL4tpGr18n9++Tu/+J1SefnbxW8mw6+50IvGxsaKuU2b\nNlXGJ06cKNYcPXp0wteu+/47d+5cZVzXtGUQ2OkCQBChCwBBhC4ABGk0m824N2s0wt5s7ty5xdzr\n16+LuYULF1bGdecc+eXvVi5+HzhwoFgzqGcY/2k2m42peN/I567O33//XczltQPv378v1ixdurQy\nrmss/88//1TGdQ006s7X8oYd/dyUYCqeu6l+5phav3vm7HQBIIjQBYAgQhcAgghdAAjSt80xWrn4\nnVJ5+dvFbzot8tdc8oK+lAb311ygF9npAkAQoQsAQYQuAATp2+YY9J5BbY5RZ/Xq1ZVxpxrL57UF\nKaV09erVSX66/qI5BtE0xwCAHiB0ASCI0AWAIEIXAIKEFlIBwCCz0wWAIEIXAIIIXQAIInQBIIjQ\nBYAgQhcAgghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIXAIIIXQAIInQBIIjQBYAgQhcA\ngghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIDMj36zRaDQj34/e0mw2G1Pxvp67wTYVz51n\nbrD97pmz0wWAIEIXAIIIXQAIInQBIIjQBYAgQhcAgghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI\n0AWAIEIXAIIIXQAIInQBIIjQBYAgM6f6AwBAL/jw4UNl/OnTp2LNunXrKuMfP35M6j3sdAEgiNAF\ngCBCFwCCCF0ACKKQCuhJEUUt9JZ58+ZVxrt37y7W5HPDw8Mde/+FCxdWxosXLy7WzJxZjU2FVADQ\no4QuAAQRugAQxJkuUGsQzteIs2nTpsr4xIkTxZq9e/dWxo1Go6ufaSI3b94s5v70GbPTBYAgQhcA\ngghdAAgidAEgiEKqaShvGpBS2TggbxqQkiIT/h9FLXTK0NBQMXfy5Mli7vjx45VxXijXrrqmKePj\n48VcXoj3118T7zkvXrzY0mtPhp0uAAQRugAQROgCQBChCwBBFFK1Ie/Uk1LZmaeue0+nuvXUFSDk\nRQJ5p56UFJ0MAkUtRKv7dzl48GBbr/Xly5fKuK7ALy8kvXv3brHm69evxdzLly8r45UrVxZrHj16\nVBk/ffr0f3/YNtnpAkAQoQsAQYQuAARxppvJmwakVJ4r5E0DUuq9xgHObweT8zW6bc2aNZVxXf1K\nnfz8/f79+8Waffv2Vcbv3r2b5Kf7186dO4u5FStWTPh3Z86cqYzfv3/f1vv/jp0uAAQRugAQROgC\nQBChCwBBBqqQqpXGAXnTgJRiGwfkTQNSaq9xgKYBg0FRC9HyJj/z589v6e9GRkYq423btnXk88ye\nPbuYy5+dlMrv0dHR0WLNw4cPO/KZfsdOFwCCCF0ACCJ0ASDIQJ3pdqpxQN40IKWycUDeNCCl1hoH\n5E0DUiobB+RNA1LSOGBQOV9jujh//nxXXvfUqVPF3ObNm4u5/Lt269atE67pBjtdAAgidAEgiNAF\ngCBCFwCC9G0hVd40IKXWGgfUNZXIGwfkTQNS6lzjgHaaBqSkcQCTo6iFdr169aoyHhsbK9bMnTu3\nmOvUv+fZs2cr47pnrs61a9cq42fPnnXk80yWnS4ABBG6ABBE6AJAkL49082bBqTUWuOAvGlASrGN\nA+p+3CBvHKBpAP9xvka0x48fV8bPnz8v1mzYsKGYW7Zs2aTf68qVK8Xc/v37K+NGo1GsuX79ejF3\n7NixSb9/N9jpAkAQoQsAQYQuAAQRugAQpG8LqdrVraYBKbXWOKCuwCVvHKBpAP9R1MJUu3XrVjFX\n98zlRaLr168v1uSFpbt27Zrw/U+fPl3MXbp0qZj7+fPnhK8VwU4XAIIIXQAIInQBIIjQBYAgfVtI\nlXfqSam1bj3d6tSTUmvdevJOPSnp1kPrFLXQq/LvxJMnTxZrli5dWhk3m81izfHjxyvjbha/doOd\nLgAEEboAEEToAkCQvj3TzZsGpNRa44B2mgakVDYOyJsGpNRa4wBNA4jgfI12DQ0NVcYbN25s6e8W\nLFgw4Zq8FmfPnj3FmgcPHrT0fr3KThcAgghdAAgidAEgiNAFgCB9W0hVp5XGAXnTgJTKxgF504CU\nOtc4QNMAJkNRC920ffv2Yu7y5cuV8fLly9t67dHR0WJuy5YtlfHIyEhbr93L7HQBIIjQBYAgQhcA\nggzUmW4r6n6oIG8ckDcNSKlsHJA3DUhJ4wD+jPM1OmnWrFnF3IULFyrjw4cPF2tmzJhRGX/79q1Y\nMz4+XswtWrSoMr5z506xZhCeMTtdAAgidAEgiNAFgCBCFwCC9G0hVd40IKXWGge00zQgpbJxgKYB\nTIaiFqINDw8Xc0eOHJnw7968eVMZ79ixo1izdu3aYu7GjRuV8apVq4o1eXOiX79+Tfh5phs7XQAI\nInQBIIjQBYAgfXOmmzcOyJsGpNS5xgF504CUnH/xZ5yvEa2u2Uqu7t88ry148uRJsabumcvVfY/m\nNQr9+MzZ6QJAEKELAEGELgAEEboAEGRaFFLljQPypgEplYf7+YF8Sq01DsibBqRUNg5QNEWnKWoh\nWivfYx8/fizmbt++XRnXNSI6dOjQhK/94sWLYm4QnjE7XQAIInQBIIjQBYAgQhcAgkyLQqq8W087\nnXpSaq1bT96pJ6WyW0/eqSelwSgAoHsUtRDt3r17xdznz58r4yVLlhRrvn//XhnXfR/OmTOnmHv7\n9m1lXPc9/vPnz/oP20fsdAEgiNAFgCBCFwCCTIsz3XYaB+RNA1LqXOOAusYbzr/4E87XiDY2NlbM\nbdq0qTI+ceJEsebo0aMTvnZdjcC5c+cq47zp0KCw0wWAIEIXAIIIXQAIInQBIEij2WzGvVmj0dab\n/f3335Vx3hAgpZTev39fGS9durRYU9c44J9//qmM6xpo5EUBebOOlBSdtKLZbDam4n3bfe6m2urV\nqyvjbha1XL16dZKfbvqYiuduuj5zdMbvnjk7XQAIInQBIIjQBYAg0+JMd+7cuZXx69evizULFy6s\njOsufrfSOCBvGpBSSgcOHKiMB/VS959ypstUcKZLNGe6ANADhC4ABBG6ABBE6AJAkGnxK0N5UVT+\nSxgplY0DWmkakFLZOCBvGpCSwikAOsNOFwCCCF0ACCJ0ASDItGiOQX/QHIOpoDkG0TTHAIAeIHQB\nIIjQBYAgQhcAgoQWUgHAILPTBYAgQhcAgghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIX\nAIIIXQAIInQBIIjQBYAgQhcAgghdAAgidAEgiNAFgCBCFwCCCF0ACCJ0ASCI0AWAIEIXAIIIXQAI\n8n8CgPonrdOClgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 576x576 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "63koWbPE5zy4",
        "colab_type": "code",
        "outputId": "9fe4e395-2621-47ad-8a3f-bed3c7378220",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "xb,yb = data.one_batch()\n",
        "xb.shape,yb.shape"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(torch.Size([128, 1, 28, 28]), torch.Size([128]))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tBb6BbUw5zy_",
        "colab_type": "code",
        "outputId": "9ff0fc5c-f6d6-485a-d912-0786ac981f01",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 369
        }
      },
      "source": [
        "data.show_batch(rows=3, figsize=(5,5))"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVIAAAFgCAYAAADpZ/FJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 360x360 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oXuof73U5zzC",
        "colab_type": "text"
      },
      "source": [
        "### Basic CNN with batchnorm"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IwArP3Ct5zzC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def conv(ni,nf): return nn.Conv2d(ni, nf, kernel_size=3, stride=2, padding=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qb8dlg8W5zzF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = nn.Sequential(\n",
        "    conv(1, 8), # 14\n",
        "    nn.BatchNorm2d(8),\n",
        "    nn.ReLU(),\n",
        "    conv(8, 16), # 7\n",
        "    nn.BatchNorm2d(16),\n",
        "    nn.ReLU(),\n",
        "    conv(16, 32), # 4\n",
        "    nn.BatchNorm2d(32),\n",
        "    nn.ReLU(),\n",
        "    conv(32, 16), # 2\n",
        "    nn.BatchNorm2d(16),\n",
        "    nn.ReLU(),\n",
        "    conv(16, 10), # 1\n",
        "    nn.BatchNorm2d(10),\n",
        "    Flatten()     # remove (1,1) grid\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jQFb5WQn5zzI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "learn = Learner(data, model, loss_func = nn.CrossEntropyLoss(), metrics=accuracy)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lyqs0eiJ5zzM",
        "colab_type": "code",
        "outputId": "fcc90ff9-7847-4271-8e34-8880e1be7219",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 765
        }
      },
      "source": [
        "print(learn.summary())"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sequential\n",
            "======================================================================\n",
            "Layer (type)         Output Shape         Param #    Trainable \n",
            "======================================================================\n",
            "Conv2d               [8, 14, 14]          80         True      \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [8, 14, 14]          16         True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [8, 14, 14]          0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 7, 7]           1,168      True      \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 7, 7]           32         True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 7, 7]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [32, 4, 4]           4,640      True      \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [32, 4, 4]           64         True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [32, 4, 4]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 2, 2]           4,624      True      \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 2, 2]           32         True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 2, 2]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [10, 1, 1]           1,450      True      \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [10, 1, 1]           20         True      \n",
            "______________________________________________________________________\n",
            "Flatten              [10]                 0          False     \n",
            "______________________________________________________________________\n",
            "\n",
            "Total params: 12,126\n",
            "Total trainable params: 12,126\n",
            "Total non-trainable params: 0\n",
            "Optimized with 'torch.optim.adam.Adam', betas=(0.9, 0.99)\n",
            "Using true weight decay as discussed in https://www.fast.ai/2018/07/02/adam-weight-decay/ \n",
            "Loss function : CrossEntropyLoss\n",
            "======================================================================\n",
            "Callbacks functions applied \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YVOpdtDF5zzR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "xb = xb.cuda()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1RaB40rr5zzT",
        "colab_type": "code",
        "outputId": "b1a491b1-bb96-4653-ce5b-244533bb37f4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "model(xb).shape"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([128, 10])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KI5_8ke-5zzW",
        "colab_type": "code",
        "outputId": "a37dd76f-1c9b-41b3-e7bb-ea00622e92fa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "learn.lr_find(end_lr=100)"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "LR Finder is complete, type {learner_name}.recorder.plot() to see the graph.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TddooCwy5zzY",
        "colab_type": "code",
        "outputId": "6ab490be-8f8c-4edc-d570-fa7eb8055c94",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        }
      },
      "source": [
        "learn.recorder.plot()"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEKCAYAAAD9xUlFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e25PECFZ5zzb",
        "colab_type": "code",
        "outputId": "93fb2ae1-580b-4776-93f6-83f3335e6be0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        }
      },
      "source": [
        "learn.fit_one_cycle(3, max_lr=0.1)"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: left;\">\n",
              "      <th>epoch</th>\n",
              "      <th>train_loss</th>\n",
              "      <th>valid_loss</th>\n",
              "      <th>accuracy</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.217328</td>\n",
              "      <td>0.412792</td>\n",
              "      <td>0.871900</td>\n",
              "      <td>00:30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.131356</td>\n",
              "      <td>0.073574</td>\n",
              "      <td>0.976300</td>\n",
              "      <td>00:29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.068040</td>\n",
              "      <td>0.040743</td>\n",
              "      <td>0.988400</td>\n",
              "      <td>00:29</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wIQNO07I5zzf",
        "colab_type": "text"
      },
      "source": [
        "### Refactor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VQSB8bdr5zzg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def conv2(ni,nf): return conv_layer(ni,nf,stride=2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EEpqntP55zzi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = nn.Sequential(\n",
        "    conv2(1, 8),   # 14\n",
        "    conv2(8, 16),  # 7\n",
        "    conv2(16, 32), # 4\n",
        "    conv2(32, 16), # 2\n",
        "    conv2(16, 10), # 1\n",
        "    Flatten()      # remove (1,1) grid\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "62jNxJX55zzk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "learn = Learner(data, model, loss_func = nn.CrossEntropyLoss(), metrics=accuracy)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VlQFX95v5zzm",
        "colab_type": "code",
        "outputId": "1e9df971-19c2-46d6-b697-00d818c26596",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        }
      },
      "source": [
        "learn.fit_one_cycle(10, max_lr=0.1)"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: left;\">\n",
              "      <th>epoch</th>\n",
              "      <th>train_loss</th>\n",
              "      <th>valid_loss</th>\n",
              "      <th>accuracy</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.234701</td>\n",
              "      <td>0.272341</td>\n",
              "      <td>0.907200</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.188814</td>\n",
              "      <td>0.146175</td>\n",
              "      <td>0.954700</td>\n",
              "      <td>00:29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.165030</td>\n",
              "      <td>0.118868</td>\n",
              "      <td>0.965500</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.142238</td>\n",
              "      <td>0.163582</td>\n",
              "      <td>0.950900</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.129594</td>\n",
              "      <td>0.083096</td>\n",
              "      <td>0.973800</td>\n",
              "      <td>00:29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.103707</td>\n",
              "      <td>0.059897</td>\n",
              "      <td>0.981100</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.083664</td>\n",
              "      <td>0.046940</td>\n",
              "      <td>0.985200</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.071521</td>\n",
              "      <td>0.041371</td>\n",
              "      <td>0.986100</td>\n",
              "      <td>00:29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.046344</td>\n",
              "      <td>0.028856</td>\n",
              "      <td>0.991400</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.045802</td>\n",
              "      <td>0.025419</td>\n",
              "      <td>0.991800</td>\n",
              "      <td>00:28</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ELyzFTEH5zzp",
        "colab_type": "text"
      },
      "source": [
        "### Resnet-ish"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u-tUcdsy5zzq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class ResBlock(nn.Module):\n",
        "    def __init__(self, nf):\n",
        "        super().__init__()\n",
        "        self.conv1 = conv_layer(nf,nf)\n",
        "        self.conv2 = conv_layer(nf,nf)\n",
        "        \n",
        "    def forward(self, x): return x + self.conv2(self.conv1(x))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W__-b1SD5zzt",
        "colab_type": "code",
        "outputId": "671da615-5cc0-44a0-f840-9258a69eba18",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "help(res_block)"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Help on function res_block in module fastai.layers:\n",
            "\n",
            "res_block(nf, dense:bool=False, norm_type:Union[fastai.layers.NormType, NoneType]=<NormType.Batch: 1>, bottle:bool=False, **conv_kwargs)\n",
            "    Resnet block of `nf` features. `conv_kwargs` are passed to `conv_layer`.\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BofJ2Cp85zz0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = nn.Sequential(\n",
        "    conv2(1, 8),\n",
        "    res_block(8),\n",
        "    conv2(8, 16),\n",
        "    res_block(16),\n",
        "    conv2(16, 32),\n",
        "    res_block(32),\n",
        "    conv2(32, 16),\n",
        "    res_block(16),\n",
        "    conv2(16, 10),\n",
        "    Flatten()\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JVdeZSaZ5zz3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def conv_and_res(ni,nf): return nn.Sequential(conv2(ni, nf), res_block(nf))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2okKwAH45zz8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = nn.Sequential(\n",
        "    conv_and_res(1, 8),\n",
        "    conv_and_res(8, 16),\n",
        "    conv_and_res(16, 32),\n",
        "    conv_and_res(32, 16),\n",
        "    conv2(16, 10),\n",
        "    Flatten()\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z_HNU-G65z0A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "learn = Learner(data, model, loss_func = nn.CrossEntropyLoss(), metrics=accuracy)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BS2coe-L5z0C",
        "colab_type": "code",
        "outputId": "37e1e2f3-4b9b-4d0c-8158-3b26f2350d66",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        }
      },
      "source": [
        "learn.lr_find(end_lr=100)\n",
        "learn.recorder.plot()"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "LR Finder is complete, type {learner_name}.recorder.plot() to see the graph.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEKCAYAAAD9xUlFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKYiISCOVgoiINFIpiIhIo/8HRA3n6GmHPyUAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-hwyudIb5z0G",
        "colab_type": "code",
        "outputId": "8d930edc-7026-4785-e06f-d8aa651215d1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 421
        }
      },
      "source": [
        "learn.fit_one_cycle(12, max_lr=0.05)"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: left;\">\n",
              "      <th>epoch</th>\n",
              "      <th>train_loss</th>\n",
              "      <th>valid_loss</th>\n",
              "      <th>accuracy</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.230527</td>\n",
              "      <td>0.167190</td>\n",
              "      <td>0.956600</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.140112</td>\n",
              "      <td>0.196680</td>\n",
              "      <td>0.934200</td>\n",
              "      <td>00:32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.102161</td>\n",
              "      <td>0.148989</td>\n",
              "      <td>0.953700</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.090911</td>\n",
              "      <td>0.164258</td>\n",
              "      <td>0.949400</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.071768</td>\n",
              "      <td>0.101622</td>\n",
              "      <td>0.967800</td>\n",
              "      <td>00:32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.065576</td>\n",
              "      <td>0.065599</td>\n",
              "      <td>0.977300</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.058586</td>\n",
              "      <td>0.080833</td>\n",
              "      <td>0.975800</td>\n",
              "      <td>00:32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.043749</td>\n",
              "      <td>0.035146</td>\n",
              "      <td>0.989100</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.039966</td>\n",
              "      <td>0.028827</td>\n",
              "      <td>0.991600</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.027281</td>\n",
              "      <td>0.019952</td>\n",
              "      <td>0.994400</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>10</td>\n",
              "      <td>0.018473</td>\n",
              "      <td>0.017786</td>\n",
              "      <td>0.994600</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>11</td>\n",
              "      <td>0.021043</td>\n",
              "      <td>0.016862</td>\n",
              "      <td>0.994800</td>\n",
              "      <td>00:31</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GjCS6hnO5z0J",
        "colab_type": "code",
        "outputId": "84ecf79d-990e-4916-fb16-4a81c84a469a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1751
        }
      },
      "source": [
        "print(learn.summary())"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sequential\n",
            "======================================================================\n",
            "Layer (type)         Output Shape         Param #    Trainable \n",
            "======================================================================\n",
            "Conv2d               [8, 14, 14]          72         True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [8, 14, 14]          0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [8, 14, 14]          16         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [8, 14, 14]          576        True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [8, 14, 14]          0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [8, 14, 14]          16         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [8, 14, 14]          576        True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [8, 14, 14]          0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [8, 14, 14]          16         True      \n",
            "______________________________________________________________________\n",
            "MergeLayer           [8, 14, 14]          0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 7, 7]           1,152      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 7, 7]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 7, 7]           32         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 7, 7]           2,304      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 7, 7]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 7, 7]           32         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 7, 7]           2,304      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 7, 7]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 7, 7]           32         True      \n",
            "______________________________________________________________________\n",
            "MergeLayer           [16, 7, 7]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [32, 4, 4]           4,608      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [32, 4, 4]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [32, 4, 4]           64         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [32, 4, 4]           9,216      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [32, 4, 4]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [32, 4, 4]           64         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [32, 4, 4]           9,216      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [32, 4, 4]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [32, 4, 4]           64         True      \n",
            "______________________________________________________________________\n",
            "MergeLayer           [32, 4, 4]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 2, 2]           4,608      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 2, 2]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 2, 2]           32         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 2, 2]           2,304      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 2, 2]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 2, 2]           32         True      \n",
            "______________________________________________________________________\n",
            "Conv2d               [16, 2, 2]           2,304      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [16, 2, 2]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [16, 2, 2]           32         True      \n",
            "______________________________________________________________________\n",
            "MergeLayer           [16, 2, 2]           0          False     \n",
            "______________________________________________________________________\n",
            "Conv2d               [10, 1, 1]           1,440      True      \n",
            "______________________________________________________________________\n",
            "ReLU                 [10, 1, 1]           0          False     \n",
            "______________________________________________________________________\n",
            "BatchNorm2d          [10, 1, 1]           20         True      \n",
            "______________________________________________________________________\n",
            "Flatten              [10]                 0          False     \n",
            "______________________________________________________________________\n",
            "\n",
            "Total params: 41,132\n",
            "Total trainable params: 41,132\n",
            "Total non-trainable params: 0\n",
            "Optimized with 'torch.optim.adam.Adam', betas=(0.9, 0.99)\n",
            "Using true weight decay as discussed in https://www.fast.ai/2018/07/02/adam-weight-decay/ \n",
            "Loss function : CrossEntropyLoss\n",
            "======================================================================\n",
            "Callbacks functions applied \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X8o4nErn5z0O",
        "colab_type": "text"
      },
      "source": [
        "## fin"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_8mpIiAg5z0P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}