fastai_imgc_plantpathology_resnet_softlabellig.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "fastai_imgc_plantpathology_resnet_softlabellig.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "koJN5rte44Ql",
        "kSAzTyj2-Yjv"
      ],
      "authorship_tag": "ABX9TyN6P5LgAEtxu7YCmj2XipHS",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "4a55b2f3f851409490cc6ce1e6b3f46e": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_7c924c797bcc44c294a16400de2cb50f",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_e6e8dd7a317540888ed88e447a223f05",
              "IPY_MODEL_2405030f86c14713adf7545dbbd6bc29"
            ]
          }
        },
        "7c924c797bcc44c294a16400de2cb50f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "e6e8dd7a317540888ed88e447a223f05": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_25203213ed854965abe14c16d66d097f",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 178728960,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 178728960,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_9ba5d905774243909a4471ff05cc44c8"
          }
        },
        "2405030f86c14713adf7545dbbd6bc29": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_beaef99c6058494abf13ee57ebc861d3",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 170M/170M [00:02&lt;00:00, 62.7MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_3f11af4bd0f34a9d969b27f2016f0337"
          }
        },
        "25203213ed854965abe14c16d66d097f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "9ba5d905774243909a4471ff05cc44c8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "beaef99c6058494abf13ee57ebc861d3": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "3f11af4bd0f34a9d969b27f2016f0337": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "c9b371e93225452fbcf7fdc33efa7ad1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_6171b207f25c41b6949da69f47830297",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_cbe3ad0ab9184f689195e898613cb689",
              "IPY_MODEL_b413c42524da464b8ba18dbddb810d3a"
            ]
          }
        },
        "6171b207f25c41b6949da69f47830297": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "cbe3ad0ab9184f689195e898613cb689": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_b3ba8369b1da4a18990a4dcd8eabf7bb",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "FloatProgressModel",
            "bar_style": "success",
            "max": 178728960,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 178728960,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_b37e94ce20a841a4b1eb21015027f2b0"
          }
        },
        "b413c42524da464b8ba18dbddb810d3a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_cee03e8681dd4291b5c196adec386de0",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 170M/170M [00:01&lt;00:00, 113MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_20e8d77cf8b44f6fa390ff95de1b9350"
          }
        },
        "b3ba8369b1da4a18990a4dcd8eabf7bb": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "b37e94ce20a841a4b1eb21015027f2b0": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "cee03e8681dd4291b5c196adec386de0": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "20e8d77cf8b44f6fa390ff95de1b9350": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/moarshy/PyTorch_fastai_Projects/blob/main/fastai_imgc_plantpathology_resnet_softlabellig.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P2zAYJZY4nFv"
      },
      "source": [
        "In this notebook softlabelling was implemented to reduce the impact of noisy labels in this [Kaggle competition](https://www.kaggle.com/c/plant-pathology-2020-fgvc7/overview)\r\n",
        "\r\n",
        "References:\r\n",
        "\r\n",
        "[notebooks](https://github.com/Isaac-Flath/kaggle/tree/main/2020-competitions/Plant_Pathology)\r\n",
        "\r\n",
        "[Isaac's Blog](https://isaac-flath.github.io/blog/deep%20learning/2021/01/18/SoftLabeling.html)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P_sADKZt4GTq",
        "outputId": "34c3721c-7bca-4e3d-a13d-e1652978c1c0"
      },
      "source": [
        "from google.colab import drive\r\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vc7CgwwU4jXr"
      },
      "source": [
        "!pip install fastai -q --upgrade"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j8ts7sPt4zgZ"
      },
      "source": [
        "from fastai.vision.all import *\r\n",
        "from sklearn.model_selection import StratifiedKFold\r\n",
        "import gc"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aYBOkyUE6vz_"
      },
      "source": [
        "SEED = 42\r\n",
        "set_seed(SEED)\r\n",
        "path = Path('/content/drive/MyDrive/colab_notebooks/fastai/plant_pathology/data')"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "koJN5rte44Ql"
      },
      "source": [
        "# PseudoLabel generation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HOdaPjPO5JHh"
      },
      "source": [
        "df = pd.read_csv(path/'train.csv')"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 200
        },
        "id": "d3lJpALB5UZW",
        "outputId": "9c5e736f-0a94-42c5-db26-224be3ed3127"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>image_id</th>\n",
              "      <th>healthy</th>\n",
              "      <th>multiple_diseases</th>\n",
              "      <th>rust</th>\n",
              "      <th>scab</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Train_0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Train_1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Train_2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Train_3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Train_4</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  image_id  healthy  multiple_diseases  rust  scab\n",
              "0  Train_0        0                  0     0     1\n",
              "1  Train_1        0                  1     0     0\n",
              "2  Train_2        1                  0     0     0\n",
              "3  Train_3        0                  0     1     0\n",
              "4  Train_4        1                  0     0     0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xPZX5CZr6fcw"
      },
      "source": [
        "df['label'] = df.iloc[:, 1:].idxmax(1)"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0jXqsrSR6RpH"
      },
      "source": [
        "#create folds\r\n",
        "N_FOLDS = 2\r\n",
        "df['fold'] = -1\r\n",
        "\r\n",
        "strat_kfold = StratifiedKFold(n_splits=N_FOLDS, random_state=SEED, shuffle=True)\r\n",
        "for i, (_, test_index) in enumerate(strat_kfold.split(df.image_id.values, df['label'].values)):\r\n",
        "    df.iloc[test_index, -1] = i\r\n",
        "    \r\n",
        "df['fold'] = df['fold'].astype('int')"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 352
        },
        "id": "J4jtpJmn60lf",
        "outputId": "e04f69f6-f732-47f8-e7ce-c2ddaa9a9c4a"
      },
      "source": [
        "df.head(10)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>image_id</th>\n",
              "      <th>healthy</th>\n",
              "      <th>multiple_diseases</th>\n",
              "      <th>rust</th>\n",
              "      <th>scab</th>\n",
              "      <th>label</th>\n",
              "      <th>fold</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Train_0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>scab</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Train_1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>multiple_diseases</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Train_2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Train_3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>rust</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Train_4</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Train_5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Train_6</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>multiple_diseases</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Train_7</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>scab</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Train_8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>scab</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Train_9</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  image_id  healthy  multiple_diseases  rust  scab              label  fold\n",
              "0  Train_0        0                  0     0     1               scab     1\n",
              "1  Train_1        0                  1     0     0  multiple_diseases     1\n",
              "2  Train_2        1                  0     0     0            healthy     0\n",
              "3  Train_3        0                  0     1     0               rust     0\n",
              "4  Train_4        1                  0     0     0            healthy     1\n",
              "5  Train_5        1                  0     0     0            healthy     0\n",
              "6  Train_6        0                  1     0     0  multiple_diseases     1\n",
              "7  Train_7        0                  0     0     1               scab     1\n",
              "8  Train_8        0                  0     0     1               scab     0\n",
              "9  Train_9        1                  0     0     0            healthy     1"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wTCDW__G5ZOi"
      },
      "source": [
        "def get_dls(df, size, fold, bs):\r\n",
        "\r\n",
        "    batch_tfms = [Normalize.from_stats(*imagenet_stats)]\r\n",
        "    dblock = DataBlock(blocks=(ImageBlock, CategoryBlock),\r\n",
        "                       splitter=IndexSplitter(df.loc[df.fold==fold].index),\r\n",
        "                       getters=[\r\n",
        "                          ColReader('image_id', pref=path/'images', suff='.jpg'),\r\n",
        "                          ColReader('label')\r\n",
        "                              ],\r\n",
        "                       item_tfms=[RandomResizedCrop(size, min_scale=0.35), FlipItem(0.5)],\r\n",
        "                       batch_tfms=batch_tfms)\r\n",
        "    return dblock.dataloaders(df, bs=bs)"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 536
        },
        "id": "xGd0L9yN7BFC",
        "outputId": "2440ec25-d172-4f6e-85a8-925586db77de"
      },
      "source": [
        "dls = get_dls(df, 224, 0, bs=16)\r\n",
        "dls.show_batch()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x648 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 810
        },
        "id": "aB1cPlua7IG2",
        "outputId": "c9e24ac6-7b03-40dd-f7f7-f67dcbb22d5e"
      },
      "source": [
        "splits, preds, targs, preds_c,  = [],[],[],[]\r\n",
        "items = pd.DataFrame(columns = df.columns)\r\n",
        "\r\n",
        "for i in range(N_FOLDS):\r\n",
        "\r\n",
        "    dls = get_dls(df, 224, i, bs=16)\r\n",
        "\r\n",
        "    learn = cnn_learner(dls, resnet18, metrics=[accuracy,RocAuc()])\r\n",
        "    learn.fine_tune(10, reset_opt=True)\r\n",
        "    \r\n",
        "    # store predictions\r\n",
        "    p, t, c = learn.get_preds(ds_idx=1, with_decoded=True)\r\n",
        "    preds.append(p); targs.append(t); preds_c.append(c); \r\n",
        "    items = pd.concat([items, dls.valid.items])"
      ],
      "execution_count": 36,
      "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.411981</td>\n",
              "      <td>0.713797</td>\n",
              "      <td>0.792536</td>\n",
              "      <td>0.870047</td>\n",
              "      <td>01:20</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.709534</td>\n",
              "      <td>0.499834</td>\n",
              "      <td>0.847420</td>\n",
              "      <td>0.911853</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.588571</td>\n",
              "      <td>0.407983</td>\n",
              "      <td>0.880351</td>\n",
              "      <td>0.935296</td>\n",
              "      <td>01:21</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.488128</td>\n",
              "      <td>0.427596</td>\n",
              "      <td>0.884742</td>\n",
              "      <td>0.938118</td>\n",
              "      <td>01:20</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.439910</td>\n",
              "      <td>0.349862</td>\n",
              "      <td>0.899012</td>\n",
              "      <td>0.947479</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.363965</td>\n",
              "      <td>0.395973</td>\n",
              "      <td>0.899012</td>\n",
              "      <td>0.941372</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.323230</td>\n",
              "      <td>0.439528</td>\n",
              "      <td>0.886937</td>\n",
              "      <td>0.931641</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.246927</td>\n",
              "      <td>0.387600</td>\n",
              "      <td>0.895719</td>\n",
              "      <td>0.939472</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.206437</td>\n",
              "      <td>0.334387</td>\n",
              "      <td>0.911087</td>\n",
              "      <td>0.945592</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.165572</td>\n",
              "      <td>0.334089</td>\n",
              "      <td>0.911087</td>\n",
              "      <td>0.947704</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.151018</td>\n",
              "      <td>0.311419</td>\n",
              "      <td>0.909989</td>\n",
              "      <td>0.951884</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.402369</td>\n",
              "      <td>0.608052</td>\n",
              "      <td>0.783516</td>\n",
              "      <td>0.894956</td>\n",
              "      <td>01:21</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.800687</td>\n",
              "      <td>0.441426</td>\n",
              "      <td>0.867033</td>\n",
              "      <td>0.927058</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.584091</td>\n",
              "      <td>0.389376</td>\n",
              "      <td>0.876923</td>\n",
              "      <td>0.943050</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.438013</td>\n",
              "      <td>0.401408</td>\n",
              "      <td>0.881319</td>\n",
              "      <td>0.945486</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.391701</td>\n",
              "      <td>0.393522</td>\n",
              "      <td>0.891209</td>\n",
              "      <td>0.948330</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.337889</td>\n",
              "      <td>0.415939</td>\n",
              "      <td>0.876923</td>\n",
              "      <td>0.951458</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.296600</td>\n",
              "      <td>0.376018</td>\n",
              "      <td>0.892308</td>\n",
              "      <td>0.959475</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.225385</td>\n",
              "      <td>0.323374</td>\n",
              "      <td>0.903297</td>\n",
              "      <td>0.966004</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.172887</td>\n",
              "      <td>0.308598</td>\n",
              "      <td>0.904396</td>\n",
              "      <td>0.964712</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.174614</td>\n",
              "      <td>0.301862</td>\n",
              "      <td>0.908791</td>\n",
              "      <td>0.964535</td>\n",
              "      <td>01:23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.175207</td>\n",
              "      <td>0.335088</td>\n",
              "      <td>0.903297</td>\n",
              "      <td>0.963807</td>\n",
              "      <td>01:22</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tS5tWTNt8oRG"
      },
      "source": [
        "imgs = L(o for o in items.image_id.values)\r\n",
        "y_true = L(o for o in items.label.values) # Labels from dataset\r\n",
        "y_targ = L(dls.vocab[o] for o in torch.cat(targs)) # Labels from out predictions\r\n",
        "y_pred = L(dls.vocab[o] for o in torch.cat(preds_c)) # predicted labels or \"pseudo labels\"\r\n",
        "p_max = torch.cat(preds).max(dim=1)[0]"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0r66x7esA81V"
      },
      "source": [
        "assert (y_true == y_targ)"
      ],
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 266
        },
        "id": "sOi1s025Bi5B",
        "outputId": "0e213d23-736a-4b79-ebcc-0f7e2947f6c7"
      },
      "source": [
        "res = pd.DataFrame({'imgs':imgs,'y_true':y_true,'y_pred':y_pred}).set_index('imgs')\r\n",
        "print(res.shape)\r\n",
        "print(df.shape)\r\n",
        "res.sample(5)"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1821, 2)\n",
            "(1821, 7)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>y_true</th>\n",
              "      <th>y_pred</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>imgs</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Train_1594</th>\n",
              "      <td>rust</td>\n",
              "      <td>rust</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Train_994</th>\n",
              "      <td>scab</td>\n",
              "      <td>scab</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Train_335</th>\n",
              "      <td>healthy</td>\n",
              "      <td>healthy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Train_1356</th>\n",
              "      <td>healthy</td>\n",
              "      <td>healthy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Train_1728</th>\n",
              "      <td>scab</td>\n",
              "      <td>scab</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "             y_true   y_pred\n",
              "imgs                        \n",
              "Train_1594     rust     rust\n",
              "Train_994      scab     scab\n",
              "Train_335   healthy  healthy\n",
              "Train_1356  healthy  healthy\n",
              "Train_1728     scab     scab"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n9oGrMnuJF_V"
      },
      "source": [
        "res.to_csv(path/'train_softlabelling.csv')"
      ],
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pBPavUa_2_xd"
      },
      "source": [
        "# Preparing functions and dataframes\r\n",
        "\r\n",
        "In this part, codes were borrowed from this excellent [blog](https://isaac-flath.github.io/blog/deep%20learning/2021/01/18/SoftLabeling.html) by Isaac"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vWyTsL2f2_JW"
      },
      "source": [
        "#df with actual and pseudolabels (from pseudolabel generation part)\r\n",
        "res = pd.read_csv(path/'train_softlabelling.csv')"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PjwhNQZvBly2"
      },
      "source": [
        "#one-hot-encode the true_label and pred_labels\r\n",
        "res = pd.get_dummies(res,columns=['y_true','y_pred']) "
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        },
        "id": "NHTQcnfyuULx",
        "outputId": "8105b520-2bab-4da5-d104-3e08fd2cadbf"
      },
      "source": [
        "res.head(3)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>imgs</th>\n",
              "      <th>y_true_healthy</th>\n",
              "      <th>y_true_multiple_diseases</th>\n",
              "      <th>y_true_rust</th>\n",
              "      <th>y_true_scab</th>\n",
              "      <th>y_pred_healthy</th>\n",
              "      <th>y_pred_multiple_diseases</th>\n",
              "      <th>y_pred_rust</th>\n",
              "      <th>y_pred_scab</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Train_2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Train_3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Train_5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      imgs  y_true_healthy  ...  y_pred_rust  y_pred_scab\n",
              "0  Train_2               1  ...            0            0\n",
              "1  Train_3               0  ...            1            0\n",
              "2  Train_5               1  ...            0            1\n",
              "\n",
              "[3 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xAAKc6QSth35"
      },
      "source": [
        "res['true_label'] = res.iloc[:, 1:5].idxmax(1).apply(lambda x: x.split('true_')[1])"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        },
        "id": "quwSHyjtByKM",
        "outputId": "c3072b84-b326-47e7-833e-ff259ba46007"
      },
      "source": [
        "res.head(3)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>imgs</th>\n",
              "      <th>y_true_healthy</th>\n",
              "      <th>y_true_multiple_diseases</th>\n",
              "      <th>y_true_rust</th>\n",
              "      <th>y_true_scab</th>\n",
              "      <th>y_pred_healthy</th>\n",
              "      <th>y_pred_multiple_diseases</th>\n",
              "      <th>y_pred_rust</th>\n",
              "      <th>y_pred_scab</th>\n",
              "      <th>true_label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Train_2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Train_3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>rust</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Train_5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>healthy</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      imgs  y_true_healthy  ...  y_pred_scab  true_label\n",
              "0  Train_2               1  ...            0     healthy\n",
              "1  Train_3               0  ...            0        rust\n",
              "2  Train_5               1  ...            1     healthy\n",
              "\n",
              "[3 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LatS4LZrFNrY"
      },
      "source": [
        "res.rename(columns={'imgs':'image_id'}, inplace=True)"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lkCeeenAJPHe"
      },
      "source": [
        "#codes taken from Isaac's blog\r\n",
        "class CrossEntropyLossOneHot(nn.Module):\r\n",
        "    def __init__(self):\r\n",
        "        super(CrossEntropyLossOneHot, self).__init__()\r\n",
        "        self.log_softmax = nn.LogSoftmax(dim=-1)\r\n",
        "\r\n",
        "    def forward(self, preds, labels):\r\n",
        "        return torch.mean(torch.sum(-labels * self.log_softmax(preds), -1))\r\n",
        "\r\n",
        "def accuracy(inp, targ, axis=-1):\r\n",
        "    \"Compute accuracy with `targ` when `pred` is bs * n_classes\"\r\n",
        "    pred,targ = flatten_check(inp.argmax(dim=axis), targ.argmax(dim=axis))\r\n",
        "    return (pred == targ).float().mean()"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sPZG99Kd3iDm"
      },
      "source": [
        "# codes taken from Isaac's blog with slight modification\r\n",
        "class SoftLabelCB(Callback):\r\n",
        "    def __init__(self, df_preds,y_true_weight = 0.7): \r\n",
        "        '''df_preds is a pandas dataframe where index is image paths\r\n",
        "             Must have y_true and y_pred one hot encoded columns (ie y_true_0, y_true_1)\r\n",
        "          '''\r\n",
        "        \r\n",
        "        self.y_true_weight = y_true_weight\r\n",
        "        self.y_pred_weight = 1 - y_true_weight\r\n",
        "        self.df = df_preds\r\n",
        "\r\n",
        "    def before_train(self):\r\n",
        "        if type(self.dl.items)==type(pd.DataFrame()): self.idx_list = L(o for o in self.dl.items.index.values)\r\n",
        "        if is_listy(self.dl.items): self.imgs_list = L(self.dl.items)      \r\n",
        "    \r\n",
        "    def before_validate(self):\r\n",
        "        if type(self.dl.items)==type(pd.DataFrame()): self.idx_list = L(o for o in self.dl.items.index.values)\r\n",
        "        if is_listy(self.dl.items): self.imgs_list = L(self.dl.items)       \r\n",
        "    \r\n",
        "    def before_batch(self):\r\n",
        "        # get the images' names for the current batch\r\n",
        "        idx = self.idx_list[self.dl._DataLoader__idxs[self.iter*self.dl.bs:self.iter*self.dl.bs+self.dl.bs]]\r\n",
        "        # get soft labels\r\n",
        "        df = self.df\r\n",
        "        soft_labels = df.loc[idx,df.columns.str.startswith('y_true')].values\r\n",
        "        \r\n",
        "        if self.training:\r\n",
        "            soft_labels = soft_labels*self.y_true_weight + df.loc[idx,df.columns.str.startswith('y_pred')].values*self.y_pred_weight\r\n",
        "        self.learn.yb = (Tensor(soft_labels).cuda(),)"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aAKs2Sjr3vpy"
      },
      "source": [
        "#create folds\r\n",
        "N_FOLDS = 3\r\n",
        "res['fold'] = -1\r\n",
        "\r\n",
        "strat_kfold = StratifiedKFold(n_splits=N_FOLDS, random_state=SEED, shuffle=True)\r\n",
        "for i, (_, test_index) in enumerate(strat_kfold.split(res.image_id.values, res.true_label.values)):\r\n",
        "    res.iloc[test_index, -1] = i\r\n",
        "    \r\n",
        "res['fold'] = res['fold'].astype('int')"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        },
        "id": "7ER8Cajm4d6G",
        "outputId": "d2efc77a-b231-45fb-8dbf-c51742c117e6"
      },
      "source": [
        "res.head(3)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>image_id</th>\n",
              "      <th>y_true_healthy</th>\n",
              "      <th>y_true_multiple_diseases</th>\n",
              "      <th>y_true_rust</th>\n",
              "      <th>y_true_scab</th>\n",
              "      <th>y_pred_healthy</th>\n",
              "      <th>y_pred_multiple_diseases</th>\n",
              "      <th>y_pred_rust</th>\n",
              "      <th>y_pred_scab</th>\n",
              "      <th>true_label</th>\n",
              "      <th>fold</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Train_2</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>healthy</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Train_3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>rust</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Train_5</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>healthy</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  image_id  y_true_healthy  ...  true_label  fold\n",
              "0  Train_2               1  ...     healthy     1\n",
              "1  Train_3               0  ...        rust     1\n",
              "2  Train_5               1  ...     healthy     2\n",
              "\n",
              "[3 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tJpKiTsU4v49",
        "outputId": "6caf16ba-d031-4e01-8a53-ba19679bc26b"
      },
      "source": [
        "res.fold.value_counts()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2    607\n",
              "1    607\n",
              "0    607\n",
              "Name: fold, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cl_TjEuz43D6"
      },
      "source": [
        "# get dls using the fold. also does oversampling of multiple_disease \r\n",
        "def get_data(fold, df):\r\n",
        "    \r\n",
        "    train_df_no_val = df[df['fold'] != {fold}]     #training set\r\n",
        "    train_df_just_val = df[df['fold'] == {fold}]   #validation set\r\n",
        "\r\n",
        "    train_df_bal = pd.concat(\r\n",
        "                   [train_df_no_val[train_df_no_val['true_label'] != \"multiple_diseases\"], train_df_just_val] +\r\n",
        "                   [train_df_no_val[train_df_no_val['true_label'] == 'multiple_diseases']] * 2\r\n",
        "                   ).sample(frac=1.0, random_state=SEED).reset_index(drop=True)\r\n",
        "    train_df_bal.reset_index(drop=True)\r\n",
        "\r\n",
        "    datablock = DataBlock(\r\n",
        "                blocks=(ImageBlock, CategoryBlock()),\r\n",
        "                getters=[\r\n",
        "                ColReader('image_id', pref=path/'images', suff='.jpg'),\r\n",
        "                ColReader('true_label')\r\n",
        "                ],\r\n",
        "                splitter=IndexSplitter(train_df_bal.loc[train_df_bal.fold==fold].index),\r\n",
        "                item_tfms=Resize(IMG_SIZE*2),\r\n",
        "                batch_tfms=[*aug_transforms(size=IMG_SIZE, max_rotate=30., min_scale=0.75, flip_vert=True, do_flip=True),\r\n",
        "                            Normalize.from_stats(*imagenet_stats)]\r\n",
        "                )\r\n",
        "    return datablock.dataloaders(source=train_df_bal, bs=BS), train_df_bal"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kSAzTyj2-Yjv"
      },
      "source": [
        "# Train without softlabelling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ifhkzrZ75mh5",
        "outputId": "95a6234d-e5cd-4bc4-a178-00fcf783cef7"
      },
      "source": [
        "IMG_SIZE = 256\r\n",
        "BS=32"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Index(['healthy', 'multiple_diseases', 'rust', 'scab'], dtype='object')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000,
          "referenced_widgets": [
            "4a55b2f3f851409490cc6ce1e6b3f46e",
            "7c924c797bcc44c294a16400de2cb50f",
            "e6e8dd7a317540888ed88e447a223f05",
            "2405030f86c14713adf7545dbbd6bc29",
            "25203213ed854965abe14c16d66d097f",
            "9ba5d905774243909a4471ff05cc44c8",
            "beaef99c6058494abf13ee57ebc861d3",
            "3f11af4bd0f34a9d969b27f2016f0337"
          ]
        },
        "id": "vITyuwX65bO6",
        "outputId": "03999e75-72f6-4c49-cd34-14e4c70abca5"
      },
      "source": [
        "for i in range(N_FOLDS):\r\n",
        "    dls = get_data(i, train_orig)\r\n",
        "    sm = SaveModelCallback(monitor='valid_loss', fname=f'{path}/model/effnet_nosl_fold_{i}_best')\r\n",
        "    learn = cnn_learner(dls, \r\n",
        "                    resnet101,\r\n",
        "                    loss_func=CrossEntropyLossFlat(),\r\n",
        "                    metrics = [error_rate, RocAuc()])\r\n",
        "    learn.fine_tune(10, reset_opt=True, cbs=sm)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading: \"https://download.pytorch.org/models/resnet101-5d3b4d8f.pth\" to /root/.cache/torch/hub/checkpoints/resnet101-5d3b4d8f.pth\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "4a55b2f3f851409490cc6ce1e6b3f46e",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=178728960.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.076039</td>\n",
              "      <td>0.861449</td>\n",
              "      <td>0.200942</td>\n",
              "      <td>0.918475</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.8614493608474731.\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.601023</td>\n",
              "      <td>0.520140</td>\n",
              "      <td>0.136578</td>\n",
              "      <td>0.953234</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.447025</td>\n",
              "      <td>0.562575</td>\n",
              "      <td>0.114600</td>\n",
              "      <td>0.962524</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.416114</td>\n",
              "      <td>0.580850</td>\n",
              "      <td>0.119309</td>\n",
              "      <td>0.950058</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.337150</td>\n",
              "      <td>0.439429</td>\n",
              "      <td>0.106750</td>\n",
              "      <td>0.968085</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.267772</td>\n",
              "      <td>0.410308</td>\n",
              "      <td>0.100471</td>\n",
              "      <td>0.967108</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.212473</td>\n",
              "      <td>0.403120</td>\n",
              "      <td>0.095761</td>\n",
              "      <td>0.968653</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.163024</td>\n",
              "      <td>0.400273</td>\n",
              "      <td>0.094192</td>\n",
              "      <td>0.963862</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.127193</td>\n",
              "      <td>0.324429</td>\n",
              "      <td>0.069074</td>\n",
              "      <td>0.971429</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.116174</td>\n",
              "      <td>0.323599</td>\n",
              "      <td>0.067504</td>\n",
              "      <td>0.972884</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.114556</td>\n",
              "      <td>0.334770</td>\n",
              "      <td>0.072214</td>\n",
              "      <td>0.973271</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.5201404690742493.\n",
            "Better model found at epoch 3 with valid_loss value: 0.43942898511886597.\n",
            "Better model found at epoch 4 with valid_loss value: 0.4103084206581116.\n",
            "Better model found at epoch 5 with valid_loss value: 0.40312016010284424.\n",
            "Better model found at epoch 6 with valid_loss value: 0.4002726972103119.\n",
            "Better model found at epoch 7 with valid_loss value: 0.324428915977478.\n",
            "Better model found at epoch 8 with valid_loss value: 0.32359927892684937.\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.086115</td>\n",
              "      <td>0.777929</td>\n",
              "      <td>0.210031</td>\n",
              "      <td>0.921712</td>\n",
              "      <td>01:33</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.7779290676116943.\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.626728</td>\n",
              "      <td>0.484421</td>\n",
              "      <td>0.128527</td>\n",
              "      <td>0.950723</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.488728</td>\n",
              "      <td>0.494639</td>\n",
              "      <td>0.115987</td>\n",
              "      <td>0.965528</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.404514</td>\n",
              "      <td>0.568086</td>\n",
              "      <td>0.103448</td>\n",
              "      <td>0.956204</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.368034</td>\n",
              "      <td>0.541154</td>\n",
              "      <td>0.109718</td>\n",
              "      <td>0.964109</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.316308</td>\n",
              "      <td>0.466378</td>\n",
              "      <td>0.114420</td>\n",
              "      <td>0.960423</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.243684</td>\n",
              "      <td>0.357282</td>\n",
              "      <td>0.094044</td>\n",
              "      <td>0.975897</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.198811</td>\n",
              "      <td>0.391327</td>\n",
              "      <td>0.087774</td>\n",
              "      <td>0.972188</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.138019</td>\n",
              "      <td>0.308911</td>\n",
              "      <td>0.081505</td>\n",
              "      <td>0.980081</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.118425</td>\n",
              "      <td>0.306226</td>\n",
              "      <td>0.078370</td>\n",
              "      <td>0.979622</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.113469</td>\n",
              "      <td>0.307329</td>\n",
              "      <td>0.079937</td>\n",
              "      <td>0.979771</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.4844209551811218.\n",
            "Better model found at epoch 4 with valid_loss value: 0.466378390789032.\n",
            "Better model found at epoch 5 with valid_loss value: 0.35728195309638977.\n",
            "Better model found at epoch 7 with valid_loss value: 0.3089105188846588.\n",
            "Better model found at epoch 8 with valid_loss value: 0.30622556805610657.\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.137273</td>\n",
              "      <td>0.780955</td>\n",
              "      <td>0.182104</td>\n",
              "      <td>0.924190</td>\n",
              "      <td>01:34</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.78095543384552.\n"
          ],
          "name": "stdout"
        },
        {
          "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>error_rate</th>\n",
              "      <th>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.675905</td>\n",
              "      <td>0.508098</td>\n",
              "      <td>0.127159</td>\n",
              "      <td>0.954580</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.501178</td>\n",
              "      <td>0.457779</td>\n",
              "      <td>0.131868</td>\n",
              "      <td>0.964304</td>\n",
              "      <td>01:34</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.423058</td>\n",
              "      <td>0.783829</td>\n",
              "      <td>0.139717</td>\n",
              "      <td>0.939096</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.336532</td>\n",
              "      <td>0.391860</td>\n",
              "      <td>0.092622</td>\n",
              "      <td>0.970471</td>\n",
              "      <td>01:34</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.299857</td>\n",
              "      <td>0.473922</td>\n",
              "      <td>0.100471</td>\n",
              "      <td>0.964449</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.235199</td>\n",
              "      <td>0.461567</td>\n",
              "      <td>0.092622</td>\n",
              "      <td>0.957917</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.195713</td>\n",
              "      <td>0.283678</td>\n",
              "      <td>0.084772</td>\n",
              "      <td>0.971444</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.153358</td>\n",
              "      <td>0.313394</td>\n",
              "      <td>0.083203</td>\n",
              "      <td>0.972161</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.109747</td>\n",
              "      <td>0.289011</td>\n",
              "      <td>0.070644</td>\n",
              "      <td>0.974289</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.095727</td>\n",
              "      <td>0.307079</td>\n",
              "      <td>0.078493</td>\n",
              "      <td>0.972947</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.5080981850624084.\n",
            "Better model found at epoch 1 with valid_loss value: 0.4577789008617401.\n",
            "Better model found at epoch 3 with valid_loss value: 0.39185968041419983.\n",
            "Better model found at epoch 6 with valid_loss value: 0.2836776375770569.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oph8CjbB51lx"
      },
      "source": [
        "#pred for submission without softlabellig\r\n",
        "test_df = pd.read_csv(path/'test.csv')\r\n",
        "test_dl = dls.test_dl(test_df)"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "id": "IewEzhnTJyIN",
        "outputId": "65115242-ffec-44cd-9ddd-28492196b537"
      },
      "source": [
        "preds = []\r\n",
        "for i in range(N_FOLDS):\r\n",
        "  learn = cnn_learner(dls, \r\n",
        "                    resnet101,\r\n",
        "                    loss_func=CrossEntropyLossFlat(),\r\n",
        "                    metrics = [error_rate, RocAuc()])\r\n",
        "  \r\n",
        "  learn.load(f'{path}/model/effnet_nosl_fold_{i}_best')\r\n",
        "\r\n",
        "  test_preds, _ = learn.get_preds(dl=test_dl)\r\n",
        "\r\n",
        "  preds.append(test_preds)"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/fastai/learner.py:56: UserWarning: Saved filed doesn't contain an optimizer state.\n",
            "  elif with_opt: warn(\"Saved filed doesn't contain an optimizer state.\")\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NKADzok5KAmF"
      },
      "source": [
        "test_preds= np.mean(np.stack(preds), axis=0)"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4Y2gg8exKE41"
      },
      "source": [
        "sub_df = pd.concat([test_df, pd.DataFrame(test_preds, columns=LABELS)], axis=1)\r\n",
        "sub_df = sub_df[['image_id', 'healthy', 'multiple_diseases', 'rust', 'scab']]\r\n",
        "sub_df.to_csv(f'{path}/submission_nosl_21012020.csv', index=False)"
      ],
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i6hs3QkDKRO-"
      },
      "source": [
        "# Train with softlabelling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R-efR9P2RI5M"
      },
      "source": [
        "IMG_SIZE = 256\r\n",
        "BS=32"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "cr-iidxDKUMd",
        "outputId": "eb9723e1-deb8-40d5-d8b7-e34939e7a19b"
      },
      "source": [
        "for i in range(N_FOLDS):\r\n",
        "    \r\n",
        "    gc.collect()\r\n",
        "\r\n",
        "    dls, df = get_data(i, res)\r\n",
        "\r\n",
        "    sm = SaveModelCallback(monitor='valid_loss', fname=f'{path}/model/effnet_sl_fold_{i}_best')\r\n",
        "\r\n",
        "    learn = cnn_learner(dls, \r\n",
        "                        resnet101,\r\n",
        "                        loss_func=CrossEntropyLossOneHot(),\r\n",
        "                        metrics = [accuracy, RocAuc()],\r\n",
        "                        cbs=SoftLabelCB(df)) #callback to softlabel the labels\r\n",
        "\r\n",
        "    learn.fine_tune(10,reset_opt=True, cbs=sm)"
      ],
      "execution_count": 133,
      "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.196069</td>\n",
              "      <td>0.646593</td>\n",
              "      <td>0.822606</td>\n",
              "      <td>0.942579</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.6465927958488464.\n"
          ],
          "name": "stdout"
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.715911</td>\n",
              "      <td>0.456973</td>\n",
              "      <td>0.869702</td>\n",
              "      <td>0.957418</td>\n",
              "      <td>01:40</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.584980</td>\n",
              "      <td>0.361855</td>\n",
              "      <td>0.908948</td>\n",
              "      <td>0.972504</td>\n",
              "      <td>01:41</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.531680</td>\n",
              "      <td>0.378312</td>\n",
              "      <td>0.902669</td>\n",
              "      <td>0.969332</td>\n",
              "      <td>01:42</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.470397</td>\n",
              "      <td>0.282871</td>\n",
              "      <td>0.919937</td>\n",
              "      <td>0.980991</td>\n",
              "      <td>01:42</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.399224</td>\n",
              "      <td>0.285282</td>\n",
              "      <td>0.923077</td>\n",
              "      <td>0.977000</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.344171</td>\n",
              "      <td>0.179385</td>\n",
              "      <td>0.943485</td>\n",
              "      <td>0.988321</td>\n",
              "      <td>01:39</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.293705</td>\n",
              "      <td>0.213124</td>\n",
              "      <td>0.924647</td>\n",
              "      <td>0.987209</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.254130</td>\n",
              "      <td>0.232186</td>\n",
              "      <td>0.924647</td>\n",
              "      <td>0.983035</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.231887</td>\n",
              "      <td>0.211991</td>\n",
              "      <td>0.927786</td>\n",
              "      <td>0.986160</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.227782</td>\n",
              "      <td>0.200714</td>\n",
              "      <td>0.937206</td>\n",
              "      <td>0.985872</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.45697343349456787.\n",
            "Better model found at epoch 1 with valid_loss value: 0.3618546426296234.\n",
            "Better model found at epoch 3 with valid_loss value: 0.28287050127983093.\n",
            "Better model found at epoch 5 with valid_loss value: 0.1793849915266037.\n"
          ],
          "name": "stdout"
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.107519</td>\n",
              "      <td>0.823140</td>\n",
              "      <td>0.816327</td>\n",
              "      <td>0.914866</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.8231399655342102.\n"
          ],
          "name": "stdout"
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.597567</td>\n",
              "      <td>0.612166</td>\n",
              "      <td>0.849294</td>\n",
              "      <td>0.939997</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.532643</td>\n",
              "      <td>0.660419</td>\n",
              "      <td>0.835165</td>\n",
              "      <td>0.955740</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.505507</td>\n",
              "      <td>0.680467</td>\n",
              "      <td>0.839874</td>\n",
              "      <td>0.930533</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.400497</td>\n",
              "      <td>0.514922</td>\n",
              "      <td>0.885400</td>\n",
              "      <td>0.955064</td>\n",
              "      <td>01:41</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.329034</td>\n",
              "      <td>0.390576</td>\n",
              "      <td>0.897959</td>\n",
              "      <td>0.967297</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.317359</td>\n",
              "      <td>0.396277</td>\n",
              "      <td>0.885400</td>\n",
              "      <td>0.956099</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.261353</td>\n",
              "      <td>0.340894</td>\n",
              "      <td>0.901099</td>\n",
              "      <td>0.967154</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.223986</td>\n",
              "      <td>0.311538</td>\n",
              "      <td>0.910518</td>\n",
              "      <td>0.971158</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.211473</td>\n",
              "      <td>0.328619</td>\n",
              "      <td>0.905808</td>\n",
              "      <td>0.970970</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.198617</td>\n",
              "      <td>0.312231</td>\n",
              "      <td>0.915228</td>\n",
              "      <td>0.972865</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.6121659874916077.\n",
            "Better model found at epoch 3 with valid_loss value: 0.5149222612380981.\n",
            "Better model found at epoch 4 with valid_loss value: 0.39057642221450806.\n",
            "Better model found at epoch 6 with valid_loss value: 0.34089395403862.\n",
            "Better model found at epoch 7 with valid_loss value: 0.3115383982658386.\n"
          ],
          "name": "stdout"
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.273679</td>\n",
              "      <td>0.828845</td>\n",
              "      <td>0.816614</td>\n",
              "      <td>0.903899</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.8288446068763733.\n"
          ],
          "name": "stdout"
        },
        {
          "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>roc_auc_score</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>0.644636</td>\n",
              "      <td>0.587642</td>\n",
              "      <td>0.869906</td>\n",
              "      <td>0.936259</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>0.538524</td>\n",
              "      <td>0.507113</td>\n",
              "      <td>0.873041</td>\n",
              "      <td>0.953461</td>\n",
              "      <td>01:38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.525435</td>\n",
              "      <td>0.561299</td>\n",
              "      <td>0.876176</td>\n",
              "      <td>0.946691</td>\n",
              "      <td>01:37</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.468210</td>\n",
              "      <td>0.399249</td>\n",
              "      <td>0.907524</td>\n",
              "      <td>0.960847</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.381069</td>\n",
              "      <td>0.295334</td>\n",
              "      <td>0.912226</td>\n",
              "      <td>0.971554</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>0.326648</td>\n",
              "      <td>0.260882</td>\n",
              "      <td>0.931035</td>\n",
              "      <td>0.974517</td>\n",
              "      <td>01:36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>6</td>\n",
              "      <td>0.295528</td>\n",
              "      <td>0.301511</td>\n",
              "      <td>0.910658</td>\n",
              "      <td>0.975747</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>7</td>\n",
              "      <td>0.253143</td>\n",
              "      <td>0.265451</td>\n",
              "      <td>0.916928</td>\n",
              "      <td>0.979405</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>8</td>\n",
              "      <td>0.233412</td>\n",
              "      <td>0.249408</td>\n",
              "      <td>0.921630</td>\n",
              "      <td>0.981532</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>9</td>\n",
              "      <td>0.219381</td>\n",
              "      <td>0.248433</td>\n",
              "      <td>0.916928</td>\n",
              "      <td>0.981433</td>\n",
              "      <td>01:35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Better model found at epoch 0 with valid_loss value: 0.5876418352127075.\n",
            "Better model found at epoch 1 with valid_loss value: 0.5071130990982056.\n",
            "Better model found at epoch 3 with valid_loss value: 0.39924871921539307.\n",
            "Better model found at epoch 4 with valid_loss value: 0.29533350467681885.\n",
            "Better model found at epoch 5 with valid_loss value: 0.2608824670314789.\n",
            "Better model found at epoch 8 with valid_loss value: 0.24940846860408783.\n",
            "Better model found at epoch 9 with valid_loss value: 0.24843280017375946.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xV5vhFiTcInh"
      },
      "source": [
        "dls, df = get_data(0, res)\r\n",
        "test_df = pd.read_csv(path/'test.csv')\r\n",
        "test_dl = dls.test_dl(test_df)"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 191,
          "referenced_widgets": [
            "c9b371e93225452fbcf7fdc33efa7ad1",
            "6171b207f25c41b6949da69f47830297",
            "cbe3ad0ab9184f689195e898613cb689",
            "b413c42524da464b8ba18dbddb810d3a",
            "b3ba8369b1da4a18990a4dcd8eabf7bb",
            "b37e94ce20a841a4b1eb21015027f2b0",
            "cee03e8681dd4291b5c196adec386de0",
            "20e8d77cf8b44f6fa390ff95de1b9350"
          ]
        },
        "id": "mntP6EaMKqdj",
        "outputId": "5734fa93-ce5c-463d-e0d6-72b114768b75"
      },
      "source": [
        "#pred for submission without softlabellig\r\n",
        "preds = []\r\n",
        "for i in range(N_FOLDS):\r\n",
        "  \r\n",
        "  learn = cnn_learner(dls, \r\n",
        "                      resnet101,\r\n",
        "                      loss_func=CrossEntropyLossOneHot(),\r\n",
        "                      metrics = [accuracy, RocAuc()],\r\n",
        "                      cbs=SoftLabelCB(df)) #callback to softlabel the labels\r\n",
        "  \r\n",
        "  learn.load(f'{path}/model/effnet_sl_fold_{i}_best')\r\n",
        "\r\n",
        "  test_preds, _ = learn.get_preds(dl=test_dl)\r\n",
        "\r\n",
        "  preds.append(test_preds)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading: \"https://download.pytorch.org/models/resnet101-5d3b4d8f.pth\" to /root/.cache/torch/hub/checkpoints/resnet101-5d3b4d8f.pth\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c9b371e93225452fbcf7fdc33efa7ad1",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(FloatProgress(value=0.0, max=178728960.0), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/fastai/learner.py:56: UserWarning: Saved filed doesn't contain an optimizer state.\n",
            "  elif with_opt: warn(\"Saved filed doesn't contain an optimizer state.\")\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/fastai/learner.py:56: UserWarning: Saved filed doesn't contain an optimizer state.\n",
            "  elif with_opt: warn(\"Saved filed doesn't contain an optimizer state.\")\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/fastai/learner.py:56: UserWarning: Saved filed doesn't contain an optimizer state.\n",
            "  elif with_opt: warn(\"Saved filed doesn't contain an optimizer state.\")\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gDsZ6_ZNeJfo"
      },
      "source": [
        "import torch.nn as nn"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M0gzHwazetzg"
      },
      "source": [
        "sm = nn.Softmax(-1)"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UGDjs5pfeFbi"
      },
      "source": [
        "p = []\r\n",
        "for i in range(3):\r\n",
        "  p.append(sm(preds[i]))"
      ],
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y6wgDnm_KwpL"
      },
      "source": [
        "test_preds= np.mean(np.stack(p), axis=0)"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yVdYvpAyKylq"
      },
      "source": [
        "sub_df = pd.concat([test_df, pd.DataFrame(test_preds, columns=['healthy', 'multiple_diseases', 'rust', 'scab'])], axis=1)\r\n",
        "sub_df = sub_df[['image_id', 'healthy', 'multiple_diseases', 'rust', 'scab']]\r\n",
        "sub_df.to_csv(f'{path}/submission_sl_21012020.csv', index=False)"
      ],
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 411
        },
        "id": "KkAtXWuRJx3j",
        "outputId": "997f1421-ab5f-47b8-9e6f-bad9279a15a5"
      },
      "source": [
        "sub_df"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>image_id</th>\n",
              "      <th>healthy</th>\n",
              "      <th>multiple_diseases</th>\n",
              "      <th>rust</th>\n",
              "      <th>scab</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Test_0</td>\n",
              "      <td>1.874524e-05</td>\n",
              "      <td>0.008526</td>\n",
              "      <td>9.914541e-01</td>\n",
              "      <td>1.090284e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Test_1</td>\n",
              "      <td>3.861142e-05</td>\n",
              "      <td>0.007045</td>\n",
              "      <td>9.926236e-01</td>\n",
              "      <td>2.930357e-04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Test_2</td>\n",
              "      <td>1.933658e-05</td>\n",
              "      <td>0.000011</td>\n",
              "      <td>1.258246e-10</td>\n",
              "      <td>9.999692e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Test_3</td>\n",
              "      <td>9.996265e-01</td>\n",
              "      <td>0.000007</td>\n",
              "      <td>3.463298e-04</td>\n",
              "      <td>1.977027e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Test_4</td>\n",
              "      <td>1.369706e-08</td>\n",
              "      <td>0.003426</td>\n",
              "      <td>9.965741e-01</td>\n",
              "      <td>8.094222e-08</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1816</th>\n",
              "      <td>Test_1816</td>\n",
              "      <td>9.570090e-08</td>\n",
              "      <td>0.000023</td>\n",
              "      <td>9.999774e-01</td>\n",
              "      <td>6.766701e-09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1817</th>\n",
              "      <td>Test_1817</td>\n",
              "      <td>5.320707e-02</td>\n",
              "      <td>0.237509</td>\n",
              "      <td>8.512319e-03</td>\n",
              "      <td>7.007719e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1818</th>\n",
              "      <td>Test_1818</td>\n",
              "      <td>2.474423e-04</td>\n",
              "      <td>0.001269</td>\n",
              "      <td>9.983528e-01</td>\n",
              "      <td>1.309485e-04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1819</th>\n",
              "      <td>Test_1819</td>\n",
              "      <td>9.882119e-01</td>\n",
              "      <td>0.000325</td>\n",
              "      <td>1.137441e-02</td>\n",
              "      <td>8.850465e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1820</th>\n",
              "      <td>Test_1820</td>\n",
              "      <td>2.579068e-03</td>\n",
              "      <td>0.112944</td>\n",
              "      <td>1.338783e-04</td>\n",
              "      <td>8.843427e-01</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1821 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       image_id       healthy  multiple_diseases          rust          scab\n",
              "0        Test_0  1.874524e-05           0.008526  9.914541e-01  1.090284e-06\n",
              "1        Test_1  3.861142e-05           0.007045  9.926236e-01  2.930357e-04\n",
              "2        Test_2  1.933658e-05           0.000011  1.258246e-10  9.999692e-01\n",
              "3        Test_3  9.996265e-01           0.000007  3.463298e-04  1.977027e-05\n",
              "4        Test_4  1.369706e-08           0.003426  9.965741e-01  8.094222e-08\n",
              "...         ...           ...                ...           ...           ...\n",
              "1816  Test_1816  9.570090e-08           0.000023  9.999774e-01  6.766701e-09\n",
              "1817  Test_1817  5.320707e-02           0.237509  8.512319e-03  7.007719e-01\n",
              "1818  Test_1818  2.474423e-04           0.001269  9.983528e-01  1.309485e-04\n",
              "1819  Test_1819  9.882119e-01           0.000325  1.137441e-02  8.850465e-05\n",
              "1820  Test_1820  2.579068e-03           0.112944  1.338783e-04  8.843427e-01\n",
              "\n",
              "[1821 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tN_l6BKTgwIj"
      },
      "source": [
        "#scores without softlabelling\r\n",
        "private score: 0.93522 public score: 0.93918\r\n",
        "\r\n",
        "#scores with softlabelling\r\n",
        "private score: 0.92873 public score: 0.93242\r\n"
      ]
    }
  ]
}