{
  "cells": [
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "HoacGdvoz75B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NpaIUZn6z8f7",
        "outputId": "7fdc9611-07d6-459e-e59a-29487c32cd69"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "vw91dJb4zsG7"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import os, gc, re, warnings\n",
        "warnings.filterwarnings(\"ignore\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "A5ZMtUPEz7Fh",
        "outputId": "253e374a-a83d-40f4-ea90-0ff5091d3e0c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                                    Text  Category\n",
              "0      बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...  Politics\n",
              "1      राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...  Politics\n",
              "2      गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...  Politics\n",
              "3      गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...  Politics\n",
              "4      गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...  Politics\n",
              "...                                                  ...       ...\n",
              "15465  । मांस काठमांडू।नेपालगुंज में लक्ष्मी अंतरसंचा...   Economy\n",
              "15466  । ३१ जुलाई काठमाण्डू।टाटा मोटर्स के आधिकारिक व...   Economy\n",
              "15467  । १३ अप्रैल काठमाण्डू।कोरियाई मोटर्स कंपनी ग्र...   Economy\n",
              "15468  । आसंजन काठमाण्डू।स्पोर्ट्स हैचबैक तसेया त्याट...   Economy\n",
              "15469  । ३ काठमाण्डू।आईएमएस मोटर्स काठमांडू के बैंक म...   Economy\n",
              "\n",
              "[15470 rows x 2 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-13cd86b6-ecb7-4c4c-8509-ad2bf2ded900\" class=\"colab-df-container\">\n",
              "    <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>Text</th>\n",
              "      <th>Category</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15465</th>\n",
              "      <td>। मांस काठमांडू।नेपालगुंज में लक्ष्मी अंतरसंचा...</td>\n",
              "      <td>Economy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15466</th>\n",
              "      <td>। ३१ जुलाई काठमाण्डू।टाटा मोटर्स के आधिकारिक व...</td>\n",
              "      <td>Economy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15467</th>\n",
              "      <td>। १३ अप्रैल काठमाण्डू।कोरियाई मोटर्स कंपनी ग्र...</td>\n",
              "      <td>Economy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15468</th>\n",
              "      <td>। आसंजन काठमाण्डू।स्पोर्ट्स हैचबैक तसेया त्याट...</td>\n",
              "      <td>Economy</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15469</th>\n",
              "      <td>। ३ काठमाण्डू।आईएमएस मोटर्स काठमांडू के बैंक म...</td>\n",
              "      <td>Economy</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>15470 rows × 2 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-13cd86b6-ecb7-4c4c-8509-ad2bf2ded900')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-13cd86b6-ecb7-4c4c-8509-ad2bf2ded900 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-13cd86b6-ecb7-4c4c-8509-ad2bf2ded900');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-9598a421-e9dc-4b3a-b870-4213601876f0\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-9598a421-e9dc-4b3a-b870-4213601876f0')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-9598a421-e9dc-4b3a-b870-4213601876f0 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "dftr = pd.read_csv(\"/content/drive/MyDrive/merged_data.csv\")\n",
        "dftr = dftr.dropna()\n",
        "dftr = dftr.reset_index(drop=True)\n",
        "dftr"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7czIovIJzRvF",
        "outputId": "c2e7b66a-a58b-4ee8-ff25-771f41fe903c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Culture          3192\n",
              "Literature       2758\n",
              "Politics         2594\n",
              "Sports           2210\n",
              "Entertainment    1078\n",
              "Opinion           992\n",
              "EduTech           784\n",
              "Economy           693\n",
              "Health            636\n",
              "Interview         533\n",
              "Name: Category, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "dftr['Category'].value_counts()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "i1-80czNzRvG"
      },
      "outputs": [],
      "source": [
        "dftr.Text = dftr.Text.astype(str)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GzYd_Boe3D9C",
        "outputId": "47e14ba8-db0c-4cde-a8de-4e83dc461af6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[K     |████████████████████████████████| 5.5 MB 15.1 MB/s \n",
            "\u001b[K     |████████████████████████████████| 182 kB 67.1 MB/s \n",
            "\u001b[K     |████████████████████████████████| 7.6 MB 54.7 MB/s \n",
            "\u001b[?25h"
          ]
        }
      ],
      "source": [
        "! pip install transformers -q"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "JskBUBgU0XY5"
      },
      "outputs": [],
      "source": [
        "from transformers import AutoModel,AutoTokenizer\n",
        "import torch\n",
        "import torch.nn.functional as F\n",
        "from tqdm import tqdm"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "htnqPsYk0YMd"
      },
      "outputs": [],
      "source": [
        "def mean_pooling(model_output, attention_mask):\n",
        "    token_embeddings = model_output.last_hidden_state.detach().cpu()\n",
        "    input_mask_expanded = (\n",
        "        attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()\n",
        "    )\n",
        "    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(\n",
        "        input_mask_expanded.sum(1), min=1e-9\n",
        "    )"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "4tpCHL360qu_"
      },
      "outputs": [],
      "source": [
        "BATCH_SIZE = 16\n",
        "\n",
        "class EmbedDataset(torch.utils.data.Dataset):\n",
        "    def __init__(self,df):\n",
        "        self.df = df.reset_index(drop=True)\n",
        "    def __len__(self):\n",
        "        return len(self.df)\n",
        "    def __getitem__(self,idx):\n",
        "        text = self.df.loc[idx,\"Text\"]\n",
        "        tokens = tokenizer(\n",
        "                text,\n",
        "                None,\n",
        "                add_special_tokens=True,\n",
        "                padding='max_length',\n",
        "                truncation=True,\n",
        "                max_length=MAX_LEN,return_tensors=\"pt\")\n",
        "        tokens = {k:v.squeeze(0) for k,v in tokens.items()}\n",
        "        return tokens\n",
        "\n",
        "ds_tr = EmbedDataset(dftr)     # <-----------------------------------------> mind your head\n",
        "#ds_tr = EmbedDataset(train)  # for emoji makes sense\n",
        "embed_dataloader_tr = torch.utils.data.DataLoader(ds_tr,\\\n",
        "                        batch_size=BATCH_SIZE,\\\n",
        "                        shuffle=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "-pN9dF9P0wpA"
      },
      "outputs": [],
      "source": [
        "tokenizer = None\n",
        "MAX_LEN = 640\n",
        "\n",
        "def get_embeddings(MODEL_NM='', MAX=640, BATCH_SIZE=4, verbose=True):\n",
        "    global tokenizer, MAX_LEN\n",
        "    DEVICE=\"cuda\"\n",
        "    model = AutoModel.from_pretrained( MODEL_NM,from_tf=True )   #TRue for TF.h5 model\n",
        "    tokenizer = AutoTokenizer.from_pretrained( MODEL_NM )\n",
        "    MAX_LEN = MAX\n",
        "\n",
        "    model = model.to(DEVICE)\n",
        "    model.eval()\n",
        "    all_train_text_feats = []\n",
        "    for batch in tqdm(embed_dataloader_tr,total=len(embed_dataloader_tr)):\n",
        "        input_ids = batch[\"input_ids\"].to(DEVICE)\n",
        "        attention_mask = batch[\"attention_mask\"].to(DEVICE)\n",
        "        with torch.no_grad():\n",
        "            model_output = model(input_ids=input_ids,attention_mask=attention_mask)\n",
        "        sentence_embeddings = mean_pooling(model_output, attention_mask.detach().cpu())\n",
        "        #sentence_embeddings = max_pooling(model_output, attention_mask.detach().cpu())    #max pooling\n",
        "        # Normalize the embeddings\n",
        "        sentence_embeddings = F.normalize(sentence_embeddings, p=2, dim=1)\n",
        "        sentence_embeddings =  sentence_embeddings.squeeze(0).detach().cpu().numpy()\n",
        "        all_train_text_feats.extend(sentence_embeddings)\n",
        "    all_train_text_feats = np.array(all_train_text_feats)\n",
        "    if verbose:\n",
        "        print('Train embeddings shape',all_train_text_feats.shape)\n",
        "\n",
        "    return all_train_text_feats"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "3KrBlZ732Qk9"
      },
      "outputs": [],
      "source": [
        "from transformers import AutoTokenizer, AutoModelForMaskedLM"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237,
          "referenced_widgets": [
            "e96896658b654eed8b968404645d0ce2",
            "18a4c50558434e0290dd7146a4fbfecf",
            "84a98a576e0249a59849cc30c7224f20",
            "e40536cd305045b0bb8aeecfd21cb956",
            "1b15bace409c4f4089f40a856b6e1338",
            "74e25a277d244040ae97abc9dde27f68",
            "eba06679f61f457b8de2817c7450ef24",
            "0b340f8a07e04471ac8ce7c17c61acba",
            "4c87386df89a4157bf67851ad066bb94",
            "f5ebefe5c7274db7b81dc329742a55e9",
            "6fc5465c7f69419ebe26ac752133676f",
            "4fcccd4ec6e54e608b823a93a59ac2b6",
            "a5ca839581e6414fba8dc778b5cf2cb9",
            "a6dbe20598484fcd9f13675e697dfc7b",
            "74f5680b28364d34a98ab4261563fc12",
            "cbdb3c1dcc984f67a5856137830afe9c",
            "4633fec68037492c92effc18055964e2",
            "3b1fbd3a31f3422aa249db2d1c933a65",
            "bd2a23d5260345279fbf87befba2e5c8",
            "6a2ed2b29770404a851a069f2a78c5da",
            "bd8042817a254f35a6ff1c52c70cead2",
            "f1a13f54dc4f48f8a613f522485755e1",
            "c40ed648d38545e390bee2431c5a707a",
            "ee7cfc940c804296894abe79e31ecbe6",
            "bc50f92e9d0d48bba21f16541fc37ed9",
            "fe96466bdfcb41739954dc0270571406",
            "69a754c07e454149aaa42ea63971f249",
            "1ca2e994552f43558ee18ea4f832d2ec",
            "b15d86eff02147bea51624bc82f94a7d",
            "ccc737848a20419684a47c4dff68dc8a",
            "aad77fce445540cfbe295df700107b6a",
            "d7bce53523fd4cafb20182edcd405533",
            "b6b326e3fe1148b29c5570b742e8b8c6"
          ]
        },
        "id": "0MTqTCpZzRvO",
        "outputId": "88d2aa13-88d2-4a11-b9a7-dd2f0fab696d"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "config.json:   0%|          | 0.00/652 [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "e96896658b654eed8b968404645d0ce2"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "tf_model.h5:   0%|          | 0.00/534M [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "4fcccd4ec6e54e608b823a93a59ac2b6"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "All TF 2.0 model weights were used when initializing BertModel.\n",
            "\n",
            "All the weights of BertModel were initialized from the TF 2.0 model.\n",
            "If your task is similar to the task the model of the checkpoint was trained on, you can already use BertModel for predictions without further training.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "vocab.txt:   0%|          | 0.00/547k [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "c40ed648d38545e390bee2431c5a707a"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 967/967 [17:21<00:00,  1.08s/it]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train embeddings shape (15470, 768)\n"
          ]
        }
      ],
      "source": [
        "MODEL_NM = 'NepBERTa/NepBERTa'\n",
        "all_train_text_feats = get_embeddings(MODEL_NM, MAX=512)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "E8yPCU5G04hq"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vmSeX8WN1ry_",
        "outputId": "d245bd98-7d63-427f-ec36-f89012ae5c87"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15470, 768)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ],
      "source": [
        "all_train_text_feats.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LeMhjb2F9q_w",
        "outputId": "caba3635-13d6-41f6-b135-aa4e31c3ce78"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "-0.028025214"
            ]
          },
          "execution_count": 47,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#all_train_text_feats[12999]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "voKcCn3k14ST"
      },
      "outputs": [],
      "source": [
        "len(all_train_text_feats)\n",
        "X_train = all_train_text_feats"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qF6rs8VS7bHg",
        "outputId": "decc601d-02b8-46e6-9177-1e27da5caf28"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15470, 768)"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ],
      "source": [
        "X_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "VdvRXyPNAZt7"
      },
      "outputs": [],
      "source": [
        "del all_train_text_feats"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "m-z0aMQOBhai"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Cverw_6bBpZs"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AtbLAHZc8Blt",
        "outputId": "1c07aed2-dfb0-41a9-89d0-bc3574d512a1"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15470,)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "y_train  = dftr['Category']\n",
        "y_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "0oDPFqEk8sE0"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "6_VI6HH_8-1n"
      },
      "outputs": [],
      "source": [
        "#y_train = y_train[:12240]\n",
        "#y_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "0nByk9Fw0up8"
      },
      "outputs": [],
      "source": [
        "Y_train = y_train\n",
        "del y_train"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "W4gcGuhS-YQ8"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "x_train, x_test, y_train, y_test = train_test_split(X_train,Y_train,test_size=0.2,random_state=42)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CI1suRpH-w8V",
        "outputId": "c3f24f19-99b7-4f91-d903-d48de3cf68b1"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((12376, 768), (12376,))"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ],
      "source": [
        "x_train.shape, y_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RPOs6N7BA84A",
        "outputId": "d0c94b27-6888-4157-ee2e-15e4c69329c5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.7307692307692306\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import LinearSVC\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.multioutput import MultiOutputClassifier\n",
        "clf=LinearSVC(penalty='l2')\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5RfaiI3O82CX"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "uy8HXUK2A9qd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d31cabd2-2af7-4544-a5ce-d9f525390e1b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.7960568842921784\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import SVC\n",
        "\n",
        "clf=SVC(kernel='rbf',gamma = 15,degree = 3, random_state=0)\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EcehuVHICpNF",
        "outputId": "4a6e0690-8f27-498f-97a4-b2b9cc061a26"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.7773109243697479\n"
          ]
        }
      ],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "lrr=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "lrr.fit(x_train, y_train)\n",
        "lrrval = lrr.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, lrrval, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AV82YWVUCxzt",
        "outputId": "1e72c563-097b-4191-b274-2e09bfd29485"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.04880413703943116\n"
          ]
        }
      ],
      "source": [
        "from sklearn.linear_model import Perceptron\n",
        "\n",
        "percept = Perceptron(alpha=1000, tol=1e-6, penalty='l2', random_state=1, max_iter=3000)\n",
        "percept = percept.fit(x_train, y_train)\n",
        "\n",
        "cpfval = percept.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, cpfval, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qK4ynXBlCp0f",
        "outputId": "e3db1c98-33e9-4a57-edb6-b6f8a9c3ef02"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.7020038784744667\n"
          ]
        }
      ],
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "clf = RandomForestClassifier(random_state=2,max_features='sqrt',class_weight='balanced')\n",
        "clf.fit(x_train, y_train)\n",
        "\n",
        "pred=clf.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, pred, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "y9wYb8D5DA4O"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rKiAyl36DB4h",
        "outputId": "a1800be3-5448-460d-cbd0-57f43d46c175"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.223792 seconds.\n",
            "You can set `force_col_wise=true` to remove the overhead.\n",
            "[LightGBM] [Info] Total Bins 195840\n",
            "[LightGBM] [Info] Number of data points in the train set: 12376, number of used features: 768\n",
            "[LightGBM] [Info] Start training from score -1.586355\n",
            "[LightGBM] [Info] Start training from score -3.128248\n",
            "[LightGBM] [Info] Start training from score -2.985763\n",
            "[LightGBM] [Info] Start training from score -2.673583\n",
            "[LightGBM] [Info] Start training from score -3.158213\n",
            "[LightGBM] [Info] Start training from score -3.332205\n",
            "[LightGBM] [Info] Start training from score -1.743339\n",
            "[LightGBM] [Info] Start training from score -2.722783\n",
            "[LightGBM] [Info] Start training from score -1.783391\n",
            "[LightGBM] [Info] Start training from score -1.930197\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] No further splits with positive gain, best gain: -inf\n",
            "[LightGBM] [Warning] Stopped training because there are no more leaves that meet the split requirements\n",
            "F1 Score : 0.31221719457013575\n"
          ]
        }
      ],
      "source": [
        "from lightgbm import LGBMClassifier\n",
        "\n",
        "model = LGBMClassifier(n_estimators = 150,learning_rate=0.5)\n",
        "model.fit(x_train, y_train)\n",
        "\n",
        "pred=model.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, pred, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 947
        },
        "id": "G7kKf9gFL4N0",
        "outputId": "affa0de7-a0d0-4232-b0de-7c6ac717b43d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.28      0.80      0.42       659\n",
            "      Economy       0.36      0.33      0.34       151\n",
            "      EduTech       0.20      0.03      0.04       159\n",
            "Entertainment       0.25      0.41      0.31       224\n",
            "       Health       0.09      0.05      0.06       110\n",
            "    Interview       0.12      0.08      0.09        91\n",
            "   Literature       0.28      0.01      0.02       593\n",
            "      Opinion       0.24      0.10      0.14       179\n",
            "     Politics       0.59      0.24      0.34       514\n",
            "       Sports       0.47      0.33      0.39       414\n",
            "\n",
            "     accuracy                           0.31      3094\n",
            "    macro avg       0.29      0.24      0.22      3094\n",
            " weighted avg       0.34      0.31      0.26      3094\n",
            "\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(y_test, pred))\n",
        "import matplotlib.pyplot as plt\n",
        "#countvector\n",
        "import seaborn as sns\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cf_matrix = confusion_matrix(y_test, pred)\n",
        "fig, ax = plt.subplots(figsize=(15,10))\n",
        "sns.heatmap(cf_matrix, linewidths=1, annot=True, ax=ax, fmt='g')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SNjAyQGcMJt0",
        "outputId": "a5a8600b-3b38-46b3-9638-675fc0ec6369"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.8093083387201034\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import SVC\n",
        "clf=SVC(kernel='rbf',gamma = 10,C = 10, random_state=42)\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 947
        },
        "id": "wF3oPRpHQy-1",
        "outputId": "10784375-2781-44d1-f380-2e7c7187e615"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.79      0.78      0.78       659\n",
            "      Economy       0.83      0.90      0.86       151\n",
            "      EduTech       0.56      0.48      0.52       159\n",
            "Entertainment       0.81      0.72      0.76       224\n",
            "       Health       0.75      0.71      0.73       110\n",
            "    Interview       0.90      0.80      0.85        91\n",
            "   Literature       0.80      0.82      0.81       593\n",
            "      Opinion       0.87      0.80      0.84       179\n",
            "     Politics       0.86      0.89      0.88       514\n",
            "       Sports       0.84      0.91      0.88       414\n",
            "\n",
            "     accuracy                           0.81      3094\n",
            "    macro avg       0.80      0.78      0.79      3094\n",
            " weighted avg       0.81      0.81      0.81      3094\n",
            "\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 31
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(y_test, clfval))\n",
        "import matplotlib.pyplot as plt\n",
        "#countvector\n",
        "import seaborn as sns\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cf_matrix = confusion_matrix(y_test, clfval)\n",
        "fig, ax = plt.subplots(figsize=(15,10))\n",
        "sns.heatmap(cf_matrix, linewidths=1, annot=True, ax=ax, fmt='g')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "id": "tMnrK8XpzRvY"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import KFold"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OJdjIRNBXyur",
        "outputId": "e0ff21e8-ef3a-45e5-dd9e-4a635c6cea66"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.8093083387201034\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.79      0.78      0.78       659\n",
            "      Economy       0.83      0.90      0.86       151\n",
            "      EduTech       0.56      0.48      0.52       159\n",
            "Entertainment       0.81      0.72      0.76       224\n",
            "       Health       0.75      0.71      0.73       110\n",
            "    Interview       0.90      0.80      0.85        91\n",
            "   Literature       0.80      0.82      0.81       593\n",
            "      Opinion       0.87      0.80      0.84       179\n",
            "     Politics       0.86      0.89      0.88       514\n",
            "       Sports       0.84      0.91      0.88       414\n",
            "\n",
            "     accuracy                           0.81      3094\n",
            "    macro avg       0.80      0.78      0.79      3094\n",
            " weighted avg       0.81      0.81      0.81      3094\n",
            "\n",
            "F1 Score : 0.826115061409179\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.79      0.84      0.81       637\n",
            "      Economy       0.77      0.80      0.78       137\n",
            "      EduTech       0.66      0.56      0.60       153\n",
            "Entertainment       0.82      0.75      0.78       221\n",
            "       Health       0.75      0.69      0.72       133\n",
            "    Interview       0.85      0.83      0.84        99\n",
            "   Literature       0.81      0.82      0.82       549\n",
            "      Opinion       0.95      0.79      0.86       206\n",
            "     Politics       0.88      0.92      0.90       496\n",
            "       Sports       0.86      0.91      0.89       463\n",
            "\n",
            "     accuracy                           0.83      3094\n",
            "    macro avg       0.82      0.79      0.80      3094\n",
            " weighted avg       0.83      0.83      0.82      3094\n",
            "\n",
            "F1 Score : 0.8306399482870072\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.81      0.81      0.81       649\n",
            "      Economy       0.84      0.83      0.83       141\n",
            "      EduTech       0.61      0.51      0.56       162\n",
            "Entertainment       0.81      0.73      0.77       183\n",
            "       Health       0.75      0.69      0.72       118\n",
            "    Interview       0.90      0.81      0.85       110\n",
            "   Literature       0.80      0.83      0.82       539\n",
            "      Opinion       0.92      0.87      0.90       212\n",
            "     Politics       0.88      0.92      0.90       550\n",
            "       Sports       0.88      0.93      0.91       430\n",
            "\n",
            "     accuracy                           0.83      3094\n",
            "    macro avg       0.82      0.79      0.81      3094\n",
            " weighted avg       0.83      0.83      0.83      3094\n",
            "\n",
            "F1 Score : 0.8180349062702004\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.78      0.83      0.81       620\n",
            "      Economy       0.82      0.86      0.84       124\n",
            "      EduTech       0.58      0.47      0.52       152\n",
            "Entertainment       0.80      0.74      0.77       211\n",
            "       Health       0.76      0.64      0.69       138\n",
            "    Interview       0.87      0.77      0.82       122\n",
            "   Literature       0.81      0.80      0.81       541\n",
            "      Opinion       0.89      0.80      0.84       207\n",
            "     Politics       0.86      0.92      0.89       535\n",
            "       Sports       0.86      0.91      0.88       444\n",
            "\n",
            "     accuracy                           0.82      3094\n",
            "    macro avg       0.80      0.78      0.79      3094\n",
            " weighted avg       0.82      0.82      0.82      3094\n",
            "\n",
            "F1 Score : 0.8112475759534583\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.78      0.79      0.79       627\n",
            "      Economy       0.87      0.80      0.83       140\n",
            "      EduTech       0.61      0.47      0.54       158\n",
            "Entertainment       0.85      0.77      0.81       239\n",
            "       Health       0.81      0.64      0.71       137\n",
            "    Interview       0.90      0.80      0.85       111\n",
            "   Literature       0.78      0.81      0.79       536\n",
            "      Opinion       0.85      0.83      0.84       188\n",
            "     Politics       0.82      0.91      0.86       499\n",
            "       Sports       0.86      0.93      0.89       459\n",
            "\n",
            "     accuracy                           0.81      3094\n",
            "    macro avg       0.81      0.77      0.79      3094\n",
            " weighted avg       0.81      0.81      0.81      3094\n",
            "\n",
            "F1 MEAN Score : 0.8190691661279896\n"
          ]
        }
      ],
      "source": [
        "kfold = KFold(n_splits=5, shuffle=True, random_state=42)\n",
        "\n",
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=SVC(kernel='rbf',gamma = 10,C = 10, random_state=42)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xOK4rYxjiM4M",
        "outputId": "9a4bc1ae-4a6c-4051-dba1-6303cba334b0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.8112475759534583\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import SVC\n",
        "clf=SVC(kernel='rbf',gamma = 10,C = 10, random_state=42)\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HXZ7CJtSiVQw",
        "outputId": "efa743ee-255c-4beb-f977-371412bc804c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.779896574014221\n"
          ]
        }
      ],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "lrr=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "lrr.fit(x_train, y_train)\n",
        "lrrval = lrr.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, lrrval, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yfDOHX5fitEa",
        "outputId": "75625486-6b76-4cd0-863d-acd66b4ed156"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "F1 Score : 0.7776341305753071\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.76      0.74      0.75       659\n",
            "      Economy       0.81      0.85      0.83       151\n",
            "      EduTech       0.51      0.40      0.45       159\n",
            "Entertainment       0.76      0.75      0.76       224\n",
            "       Health       0.71      0.67      0.69       110\n",
            "    Interview       0.78      0.74      0.76        91\n",
            "   Literature       0.76      0.80      0.78       593\n",
            "      Opinion       0.76      0.73      0.75       179\n",
            "     Politics       0.86      0.89      0.87       514\n",
            "       Sports       0.83      0.86      0.84       414\n",
            "\n",
            "     accuracy                           0.78      3094\n",
            "    macro avg       0.75      0.74      0.75      3094\n",
            " weighted avg       0.77      0.78      0.78      3094\n",
            "\n",
            "F1 Score : 0.8002585649644475\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.78      0.80      0.79       637\n",
            "      Economy       0.73      0.80      0.76       137\n",
            "      EduTech       0.61      0.47      0.53       153\n",
            "Entertainment       0.78      0.74      0.76       221\n",
            "       Health       0.75      0.67      0.71       133\n",
            "    Interview       0.75      0.73      0.74        99\n",
            "   Literature       0.77      0.80      0.78       549\n",
            "      Opinion       0.89      0.74      0.81       206\n",
            "     Politics       0.87      0.92      0.89       496\n",
            "       Sports       0.85      0.89      0.87       463\n",
            "\n",
            "     accuracy                           0.80      3094\n",
            "    macro avg       0.78      0.76      0.77      3094\n",
            " weighted avg       0.80      0.80      0.80      3094\n",
            "\n",
            "F1 Score : 0.7892695539754362\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.78      0.75      0.76       649\n",
            "      Economy       0.82      0.83      0.83       141\n",
            "      EduTech       0.58      0.40      0.47       162\n",
            "Entertainment       0.71      0.73      0.72       183\n",
            "       Health       0.73      0.64      0.68       118\n",
            "    Interview       0.80      0.75      0.77       110\n",
            "   Literature       0.73      0.81      0.76       539\n",
            "      Opinion       0.88      0.77      0.82       212\n",
            "     Politics       0.85      0.91      0.88       550\n",
            "       Sports       0.85      0.91      0.88       430\n",
            "\n",
            "     accuracy                           0.79      3094\n",
            "    macro avg       0.77      0.75      0.76      3094\n",
            " weighted avg       0.79      0.79      0.79      3094\n",
            "\n",
            "F1 Score : 0.7805429864253394\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.75      0.79      0.77       620\n",
            "      Economy       0.77      0.80      0.79       124\n",
            "      EduTech       0.50      0.35      0.41       152\n",
            "Entertainment       0.76      0.72      0.74       211\n",
            "       Health       0.73      0.57      0.64       138\n",
            "    Interview       0.88      0.71      0.79       122\n",
            "   Literature       0.74      0.77      0.76       541\n",
            "      Opinion       0.79      0.74      0.77       207\n",
            "     Politics       0.85      0.90      0.88       535\n",
            "       Sports       0.84      0.91      0.87       444\n",
            "\n",
            "     accuracy                           0.78      3094\n",
            "    macro avg       0.76      0.73      0.74      3094\n",
            " weighted avg       0.78      0.78      0.78      3094\n",
            "\n",
            "F1 Score : 0.7802197802197802\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "      Culture       0.77      0.74      0.76       627\n",
            "      Economy       0.79      0.74      0.76       140\n",
            "      EduTech       0.52      0.36      0.43       158\n",
            "Entertainment       0.78      0.76      0.77       239\n",
            "       Health       0.78      0.61      0.69       137\n",
            "    Interview       0.88      0.78      0.83       111\n",
            "   Literature       0.73      0.79      0.76       536\n",
            "      Opinion       0.82      0.81      0.81       188\n",
            "     Politics       0.81      0.89      0.85       499\n",
            "       Sports       0.84      0.90      0.87       459\n",
            "\n",
            "     accuracy                           0.78      3094\n",
            "    macro avg       0.77      0.74      0.75      3094\n",
            " weighted avg       0.78      0.78      0.78      3094\n",
            "\n",
            "F1 MEAN Score : 0.785585003232062\n"
          ]
        }
      ],
      "source": [
        "kfold = KFold(n_splits=5, shuffle=True, random_state=42)\n",
        "\n",
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 652
        },
        "id": "wTSCOpD9kXYA",
        "outputId": "aaf2cace-68c9-4a0d-c344-a3d104782321"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 37
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cf_matrix = confusion_matrix(y_test, clfval)\n",
        "fig, ax = plt.subplots(figsize=(15,10))\n",
        "sns.heatmap(cf_matrix, linewidths=1, annot=True, ax=ax, fmt='g')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iM8dSpKqZV-h"
      },
      "outputs": [],
      "source": [
        "# nepali BERT"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255,
          "referenced_widgets": [
            "dd083e9312f34cfe8712574bd16c2520",
            "54891ad613ed4f27bb3aa9be849132b1",
            "007304dd7fb74a1b9d3b390f13e73a03",
            "4298e20b35474077a0f872ddbc6db987",
            "076956eb941c4cd6a402c1335aa3bba1",
            "4b7ba118f59646e59cec9056ffcbcaa3",
            "525d941c1e3541919bacb96f9adfdd91",
            "821aa443392a4ac99d1adb7cc9138840",
            "888b6d0488a649f7a4469b93f1bc7bc6",
            "20b7ae08f2de4b86bda6c5f94054eb75",
            "879f999eba754a20861a37fcb275c7e0",
            "fe72571856854a17bf5df0b4ff3ff604",
            "d7c45455a8e8445c929c8c83ab1f9fa1",
            "748462ab70364e168c082c7573b52238",
            "67fdea3f80f54b849854ec829e37d80b",
            "7a1044e4ca444ae990d163dff600a113",
            "f58360c8802b490fb23cb4a17a81911e",
            "a42f03dc2a5849beb94666ff1648ff5b",
            "020a891f78e14676b691abd046356725",
            "0550c59aa57f40e094beae25cc5b0dc3",
            "118705ff5aa84e59801ce43e13bf782f",
            "23531ca8f9e14e7dbc8f1ab522131142",
            "4528dd10b894401e98dcaf0dd547bf13",
            "d7892059dc6048b0bfa8e18998125727",
            "8664160808b0488bba20727074611c9f",
            "e9f2d971371a4c5cac7c9eab4ec30d89",
            "511fdf809305494ca560dd7a84e8f904",
            "802d3e98e6b94a6881d3eb2f11ddb235",
            "a360588788c34e59869144df00b459ec",
            "a8395f7eabf54ffc8329dfed96cc400a",
            "2f56fa2517b2460a8d432ea866b657dd",
            "5253b2271568488da290f23826fa8979",
            "ae16930d4b7e498e888c0b1a26ca9ecc"
          ]
        },
        "id": "UlN39Ni9ZVxc",
        "outputId": "9076dee9-3bde-4651-b872-09487e2e77fc"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "dd083e9312f34cfe8712574bd16c2520",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/589 [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "fe72571856854a17bf5df0b4ff3ff604",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/438M [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Some weights of the model checkpoint at Shushant/nepaliBERT were not used when initializing BertModel: ['cls.predictions.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.decoder.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.decoder.weight']\n",
            "- This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
            "- This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
            "Some weights of BertModel were not initialized from the model checkpoint at Shushant/nepaliBERT and are newly initialized: ['bert.pooler.dense.weight', 'bert.pooler.dense.bias']\n",
            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "4528dd10b894401e98dcaf0dd547bf13",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/529k [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 1531/1531 [07:06<00:00,  3.59it/s]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Train embeddings shape (13008,)\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        }
      ],
      "source": [
        "MODEL_NM = 'Shushant/nepaliBERT'\n",
        "all_train_text_feats = get_embeddings(MODEL_NM, MAX=512)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uTZHO-YKZZz8",
        "outputId": "2e5f0141-83f9-47ad-c14c-879e24b7fe77"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(12240, 768)"
            ]
          },
          "execution_count": 11,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X_train = all_train_text_feats[:12240]\n",
        "\n",
        "X_tr = []\n",
        "for arr in X_train:\n",
        "  X_tr.append(arr.reshape(768))\n",
        "\n",
        "X_train = np.array(X_tr)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t9Ghp97gbaRH",
        "outputId": "54576dfa-4def-4bec-db48-aaafdb33d299"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "((12240, 768), (12240,))"
            ]
          },
          "execution_count": 16,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "Y_train = y_train\n",
        "X_train.shape, Y_train.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "53kSPyWPbmGC",
        "outputId": "37d6bfec-eda2-4a0d-8e1d-d0713e98bca0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.5870098039215687\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.70      0.61      0.65       604\n",
            "           1       0.52      0.97      0.68       809\n",
            "           2       0.15      0.05      0.08       173\n",
            "           3       0.64      0.34      0.44       219\n",
            "           4       0.89      0.56      0.68       234\n",
            "           5       0.61      0.14      0.22       148\n",
            "           6       0.58      0.06      0.11       118\n",
            "           7       0.87      0.34      0.48       143\n",
            "\n",
            "    accuracy                           0.59      2448\n",
            "   macro avg       0.62      0.38      0.42      2448\n",
            "weighted avg       0.61      0.59      0.54      2448\n",
            "\n",
            "F1 Score : 0.6491013071895425\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.71      0.82      0.76       604\n",
            "           1       0.61      0.98      0.75       809\n",
            "           2       0.68      0.16      0.25       174\n",
            "           3       0.62      0.38      0.47       218\n",
            "           4       0.80      0.56      0.66       234\n",
            "           5       0.38      0.08      0.13       148\n",
            "           6       0.29      0.05      0.09       118\n",
            "           7       0.77      0.34      0.47       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.61      0.42      0.45      2448\n",
            "weighted avg       0.64      0.65      0.60      2448\n",
            "\n",
            "F1 Score : 0.6580882352941176\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.74      0.80      0.77       604\n",
            "           1       0.61      0.97      0.75       808\n",
            "           2       0.55      0.13      0.21       174\n",
            "           3       0.64      0.47      0.54       218\n",
            "           4       0.77      0.55      0.64       234\n",
            "           5       0.50      0.14      0.22       149\n",
            "           6       0.83      0.17      0.28       118\n",
            "           7       0.71      0.33      0.45       143\n",
            "\n",
            "    accuracy                           0.66      2448\n",
            "   macro avg       0.67      0.45      0.48      2448\n",
            "weighted avg       0.66      0.66      0.61      2448\n",
            "\n",
            "F1 Score : 0.6070261437908496\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.75      0.62      0.68       604\n",
            "           1       0.56      0.95      0.71       808\n",
            "           2       0.34      0.09      0.15       173\n",
            "           3       0.58      0.47      0.52       219\n",
            "           4       0.67      0.62      0.64       233\n",
            "           5       0.50      0.17      0.25       149\n",
            "           6       0.29      0.06      0.10       119\n",
            "           7       0.75      0.29      0.42       143\n",
            "\n",
            "    accuracy                           0.61      2448\n",
            "   macro avg       0.55      0.41      0.43      2448\n",
            "weighted avg       0.60      0.61      0.56      2448\n",
            "\n",
            "F1 Score : 0.5812908496732027\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.79      0.50      0.61       604\n",
            "           1       0.51      0.98      0.67       808\n",
            "           2       0.57      0.15      0.24       173\n",
            "           3       0.57      0.41      0.47       219\n",
            "           4       0.81      0.50      0.62       233\n",
            "           5       0.46      0.17      0.25       149\n",
            "           6       0.52      0.09      0.16       118\n",
            "           7       0.61      0.42      0.49       144\n",
            "\n",
            "    accuracy                           0.58      2448\n",
            "   macro avg       0.61      0.40      0.44      2448\n",
            "weighted avg       0.62      0.58      0.54      2448\n",
            "\n",
            "F1 MEAN Score : 0.6165032679738562\n"
          ]
        }
      ],
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.model_selection import StratifiedKFold\n",
        "kfold = StratifiedKFold(n_splits=5)\n",
        "\n",
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=SVC(kernel='rbf',gamma = 10,C = 10, random_state=42)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MItgmLuTb-Im",
        "outputId": "f310d14a-ee29-4962-8226-e590aafb2929"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6327614379084967\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.64      0.73      0.68       604\n",
            "           1       0.61      0.95      0.74       809\n",
            "           2       0.12      0.03      0.05       173\n",
            "           3       0.61      0.47      0.53       219\n",
            "           4       0.87      0.66      0.75       234\n",
            "           5       0.50      0.07      0.13       148\n",
            "           6       0.71      0.08      0.15       118\n",
            "           7       0.79      0.36      0.50       143\n",
            "\n",
            "    accuracy                           0.63      2448\n",
            "   macro avg       0.61      0.42      0.44      2448\n",
            "weighted avg       0.62      0.63      0.58      2448\n",
            "\n",
            "F1 Score : 0.6736111111111112\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.68      0.87      0.76       604\n",
            "           1       0.67      0.98      0.80       809\n",
            "           2       0.62      0.07      0.13       174\n",
            "           3       0.58      0.49      0.53       218\n",
            "           4       0.79      0.71      0.75       234\n",
            "           5       0.25      0.03      0.06       148\n",
            "           6       0.44      0.06      0.10       118\n",
            "           7       0.71      0.28      0.40       143\n",
            "\n",
            "    accuracy                           0.67      2448\n",
            "   macro avg       0.59      0.44      0.44      2448\n",
            "weighted avg       0.64      0.67      0.61      2448\n",
            "\n",
            "F1 Score : 0.6858660130718954\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.70      0.83      0.76       604\n",
            "           1       0.67      0.96      0.79       808\n",
            "           2       0.27      0.04      0.07       174\n",
            "           3       0.62      0.60      0.61       218\n",
            "           4       0.79      0.75      0.77       234\n",
            "           5       0.62      0.11      0.18       149\n",
            "           6       0.84      0.18      0.29       118\n",
            "           7       0.75      0.36      0.48       143\n",
            "\n",
            "    accuracy                           0.69      2448\n",
            "   macro avg       0.66      0.48      0.49      2448\n",
            "weighted avg       0.67      0.69      0.63      2448\n",
            "\n",
            "F1 Score : 0.6462418300653595\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.71      0.69      0.70       604\n",
            "           1       0.65      0.96      0.78       808\n",
            "           2       0.50      0.09      0.16       173\n",
            "           3       0.47      0.52      0.50       219\n",
            "           4       0.66      0.82      0.73       233\n",
            "           5       0.54      0.13      0.21       149\n",
            "           6       0.46      0.09      0.15       119\n",
            "           7       0.73      0.31      0.43       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.59      0.45      0.46      2448\n",
            "weighted avg       0.63      0.65      0.60      2448\n",
            "\n",
            "F1 Score : 0.6029411764705882\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.76      0.58      0.66       604\n",
            "           1       0.56      0.97      0.71       808\n",
            "           2       0.31      0.03      0.05       173\n",
            "           3       0.46      0.45      0.46       219\n",
            "           4       0.82      0.62      0.70       233\n",
            "           5       0.47      0.10      0.17       149\n",
            "           6       0.50      0.08      0.13       118\n",
            "           7       0.57      0.47      0.52       144\n",
            "\n",
            "    accuracy                           0.60      2448\n",
            "   macro avg       0.56      0.41      0.42      2448\n",
            "weighted avg       0.60      0.60      0.55      2448\n",
            "\n",
            "F1 MEAN Score : 0.6482843137254902\n"
          ]
        }
      ],
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=RandomForestClassifier()\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "70f5YoVogSl2",
        "outputId": "1e35e0fd-6329-4545-b568-5b994ff285cc"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6907679738562091\n"
          ]
        }
      ],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "lrr=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "lrr.fit(x_train, y_train)\n",
        "lrrval = lrr.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, lrrval, average='micro')))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rgmTL-nLfjtF",
        "outputId": "738fcaf9-acf8-4747-9ac7-801c27dbe5ea"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6895424836601307\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.68      0.72      0.70       604\n",
            "           1       0.86      0.89      0.88       809\n",
            "           2       0.41      0.51      0.45       173\n",
            "           3       0.62      0.53      0.57       219\n",
            "           4       0.84      0.70      0.76       234\n",
            "           5       0.37      0.33      0.35       148\n",
            "           6       0.26      0.26      0.26       118\n",
            "           7       0.70      0.61      0.65       143\n",
            "\n",
            "    accuracy                           0.69      2448\n",
            "   macro avg       0.59      0.57      0.58      2448\n",
            "weighted avg       0.69      0.69      0.69      2448\n",
            "\n",
            "F1 Score : 0.7258986928104575\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.71      0.82      0.76       604\n",
            "           1       0.90      0.92      0.91       809\n",
            "           2       0.41      0.38      0.40       174\n",
            "           3       0.62      0.54      0.57       218\n",
            "           4       0.79      0.80      0.80       234\n",
            "           5       0.40      0.33      0.36       148\n",
            "           6       0.37      0.31      0.33       118\n",
            "           7       0.66      0.57      0.61       143\n",
            "\n",
            "    accuracy                           0.73      2448\n",
            "   macro avg       0.61      0.58      0.59      2448\n",
            "weighted avg       0.72      0.73      0.72      2448\n",
            "\n",
            "F1 Score : 0.7475490196078431\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.75      0.84      0.79       604\n",
            "           1       0.91      0.91      0.91       808\n",
            "           2       0.43      0.35      0.39       174\n",
            "           3       0.65      0.70      0.67       218\n",
            "           4       0.81      0.78      0.79       234\n",
            "           5       0.46      0.38      0.41       149\n",
            "           6       0.50      0.39      0.44       118\n",
            "           7       0.64      0.64      0.64       143\n",
            "\n",
            "    accuracy                           0.75      2448\n",
            "   macro avg       0.64      0.62      0.63      2448\n",
            "weighted avg       0.74      0.75      0.74      2448\n",
            "\n",
            "F1 Score : 0.6895424836601307\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.81      0.71      0.76       604\n",
            "           1       0.90      0.85      0.87       808\n",
            "           2       0.36      0.29      0.32       173\n",
            "           3       0.50      0.61      0.55       219\n",
            "           4       0.63      0.83      0.72       233\n",
            "           5       0.35      0.38      0.37       149\n",
            "           6       0.37      0.46      0.41       119\n",
            "           7       0.63      0.55      0.59       143\n",
            "\n",
            "    accuracy                           0.69      2448\n",
            "   macro avg       0.57      0.59      0.57      2448\n",
            "weighted avg       0.70      0.69      0.69      2448\n",
            "\n",
            "F1 Score : 0.6646241830065359\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.78      0.51      0.62       604\n",
            "           1       0.85      0.90      0.88       808\n",
            "           2       0.37      0.40      0.38       173\n",
            "           3       0.47      0.60      0.53       219\n",
            "           4       0.79      0.72      0.75       233\n",
            "           5       0.36      0.42      0.39       149\n",
            "           6       0.43      0.36      0.40       118\n",
            "           7       0.47      0.78      0.59       144\n",
            "\n",
            "    accuracy                           0.66      2448\n",
            "   macro avg       0.57      0.59      0.57      2448\n",
            "weighted avg       0.69      0.66      0.67      2448\n",
            "\n",
            "F1 MEAN Score : 0.7034313725490197\n"
          ]
        }
      ],
      "source": [
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I3A9KryRgOAE",
        "outputId": "79ff225e-5de8-4bbe-b97a-2d33e276b2ad"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6997549019607843\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import SVC\n",
        "clf=SVC(kernel='rbf',degree = 3, random_state=0,max_iter=10000)\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8_Vqzbq6jZTM",
        "outputId": "da452467-029e-4268-a344-5f966e0782ac"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6703431372549019\n"
          ]
        }
      ],
      "source": [
        "from lightgbm import LGBMClassifier\n",
        "\n",
        "model = LGBMClassifier()\n",
        "model.fit(x_train, y_train)\n",
        "\n",
        "pred=model.predict(x_test)\n",
        "print('F1 Score : {}'.format(f1_score(y_test, pred, average='micro')))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uTGtfyMelC-Y"
      },
      "source": [
        "# deberta base nepali"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284,
          "referenced_widgets": [
            "b1715cd48f6741088b3bf7b22f21030f",
            "3d7fc90750fc4fd28b2b6bef0ca6e601",
            "fcd6ab186c394b4fab091e2ec64cf7d2",
            "70d42517af9f49e38cf99b7c97b05239",
            "8445fd819a6a46fcbf75e99cbb18b4ce",
            "0ddf0870774544bc9122761b7968c2ec",
            "68e41c648aef493ab2290e9bd95ab244",
            "95400b69037c4c3b8ce7ada698d046ce",
            "d0979c33cdc54526b5dd002b2c23bef6",
            "d3b1b8f83ae34e7d90ab862e23d2d5d9",
            "9afec31fa2c542edbc191282ee03df73",
            "4d8556c2c15046f481701f853e13b126",
            "8c3c9b20a3154dceb842acbf88160743",
            "7dc6e325b5454fe294d3fe960fa8f2f0",
            "363c84ef86a34e6a93f4db2723a3c9f4",
            "d96a8fd5fb1540a0902e1fac49912602",
            "18c49519429c41f195d703bed69a4443",
            "36620b6e478042989fe2ce175b627125",
            "708178f346da4f1597853c0218a73e03",
            "c792e1ae15ce4fa29ccac9e75780bdd1",
            "768426a2c2ee437389cdded02f3f0c48",
            "44323c2a91184648aa0af06e91ab980d",
            "a2b590b036264886835d76ebae8de084",
            "5d7a2ed7e1eb431ca509ede35d4e562f",
            "714bdef5435d45ac8dbb7666c0e7232e",
            "32293f1a4a3e4b49b27f4be3fe5c3b1a",
            "0766b7d42af943d0b4d37cab4bddf569",
            "0c31a369e5db4e0bbf689734f4dcd415",
            "ec44446c5ce541e4a23d601aaa581a69",
            "7ceed10a595343168a068bebf9746bca",
            "67c61a63e12e48dda06d76fd11a0d7e5",
            "bc565a24c7fb4bb799eb45d251cc9a57",
            "14feea72286044c3ac19c79bdb4a5d15",
            "c100b12a53de481ea1e5bce81995319c",
            "499f1d9965704b1a9601ca6701e886de",
            "662f45e53d6d45f49c249cabc6f1dc91",
            "51bbf362fa7a49d0802a970b5443b01d",
            "1a602627ce7742e7bf9af421ab365430",
            "9541cd9ae94e469f8c0ed5ce66a0b01d",
            "c4dbec2e026a4722bbf17bef4ad5741b",
            "a79e620eb592460e986ea5c305ebeaec",
            "39a62a0780dc4911bd30d366ac81b6f3",
            "f183c27d35414a5facef4bed84aded9b",
            "3367b5295e4b44d5a3aee0aaf5fbd73a",
            "dca828d0f4824d5a80125caf02b1bd18",
            "3cdaa669bed14105a18bb904190b43a6",
            "9dfcb1560dcd483ab5c000d6e43771f3",
            "91b65e18dfae4ee08c23d8d70436f3d2",
            "a054cd75df6a40a1a8a23cb82c73e73d",
            "c9a5a153a6d8456fb867a2bbeee2368d",
            "7d940db073774c888ab9a98316fbc3ae",
            "b5422f0f1b65405ea399b64dd602c2a4",
            "93fa5389c5e844d98eb849c08c035c1d",
            "6ebab858b53d466a96db54ced3956fd0",
            "1967dfefa2e24c76a680a450dfa66e6f"
          ]
        },
        "id": "mpZVagj6jgJf",
        "outputId": "45f9e81f-c0d1-4ffd-891f-dea9455b06eb"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "b1715cd48f6741088b3bf7b22f21030f",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/773 [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "4d8556c2c15046f481701f853e13b126",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/557M [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Some weights of the model checkpoint at Sakonii/deberta-base-nepali were not used when initializing DebertaModel: ['cls.predictions.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.decoder.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.decoder.weight']\n",
            "- This IS expected if you are initializing DebertaModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
            "- This IS NOT expected if you are initializing DebertaModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "a2b590b036264886835d76ebae8de084",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/413 [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c100b12a53de481ea1e5bce81995319c",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/2.13M [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "dca828d0f4824d5a80125caf02b1bd18",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "Downloading:   0%|          | 0.00/239 [00:00<?, ?B/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 1531/1531 [10:41<00:00,  2.38it/s]"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Train embeddings shape (13008,)\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        }
      ],
      "source": [
        "MODEL_NM = 'Sakonii/deberta-base-nepali'\n",
        "all_train_text_feats = get_embeddings(MODEL_NM, MAX=512)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WZf17q46lKf1"
      },
      "outputs": [],
      "source": [
        "X_train = all_train_text_feats[:12240]\n",
        "\n",
        "X_tr = []\n",
        "for arr in X_train:\n",
        "  X_tr.append(arr.reshape(768))\n",
        "\n",
        "X_train = np.array(X_tr)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C8cvYO5Unsq_",
        "outputId": "7ff8a0ed-7f42-4d16-93fa-088f669c1c4d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6429738562091504\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.61      0.72      0.66       604\n",
            "           1       0.67      0.93      0.78       809\n",
            "           2       0.24      0.17      0.20       173\n",
            "           3       0.69      0.58      0.63       219\n",
            "           4       0.87      0.53      0.66       234\n",
            "           5       0.50      0.26      0.34       148\n",
            "           6       0.44      0.13      0.20       118\n",
            "           7       0.81      0.43      0.56       143\n",
            "\n",
            "    accuracy                           0.64      2448\n",
            "   macro avg       0.60      0.47      0.50      2448\n",
            "weighted avg       0.63      0.64      0.62      2448\n",
            "\n",
            "F1 Score : 0.6879084967320261\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.64      0.86      0.74       604\n",
            "           1       0.78      0.93      0.84       809\n",
            "           2       0.45      0.17      0.25       174\n",
            "           3       0.59      0.57      0.58       218\n",
            "           4       0.84      0.67      0.75       234\n",
            "           5       0.39      0.24      0.30       148\n",
            "           6       0.31      0.09      0.14       118\n",
            "           7       0.66      0.39      0.49       143\n",
            "\n",
            "    accuracy                           0.69      2448\n",
            "   macro avg       0.58      0.49      0.51      2448\n",
            "weighted avg       0.66      0.69      0.66      2448\n",
            "\n",
            "F1 Score : 0.7156862745098039\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.70      0.85      0.77       604\n",
            "           1       0.78      0.93      0.85       808\n",
            "           2       0.55      0.25      0.35       174\n",
            "           3       0.62      0.71      0.66       218\n",
            "           4       0.79      0.69      0.73       234\n",
            "           5       0.51      0.32      0.39       149\n",
            "           6       0.67      0.24      0.35       118\n",
            "           7       0.68      0.40      0.50       143\n",
            "\n",
            "    accuracy                           0.72      2448\n",
            "   macro avg       0.66      0.55      0.57      2448\n",
            "weighted avg       0.70      0.72      0.69      2448\n",
            "\n",
            "F1 Score : 0.6544117647058824\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.69      0.67      0.68       604\n",
            "           1       0.72      0.90      0.80       808\n",
            "           2       0.35      0.17      0.23       173\n",
            "           3       0.57      0.61      0.59       219\n",
            "           4       0.63      0.79      0.70       233\n",
            "           5       0.48      0.39      0.43       149\n",
            "           6       0.46      0.18      0.26       119\n",
            "           7       0.60      0.31      0.41       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.56      0.50      0.51      2448\n",
            "weighted avg       0.63      0.65      0.63      2448\n",
            "\n",
            "F1 Score : 0.6625816993464052\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.74      0.59      0.65       604\n",
            "           1       0.71      0.93      0.80       808\n",
            "           2       0.48      0.31      0.37       173\n",
            "           3       0.53      0.71      0.61       219\n",
            "           4       0.85      0.61      0.71       233\n",
            "           5       0.42      0.34      0.37       149\n",
            "           6       0.45      0.14      0.22       118\n",
            "           7       0.56      0.69      0.62       144\n",
            "\n",
            "    accuracy                           0.66      2448\n",
            "   macro avg       0.59      0.54      0.54      2448\n",
            "weighted avg       0.66      0.66      0.64      2448\n",
            "\n",
            "F1 MEAN Score : 0.6727124183006536\n"
          ]
        }
      ],
      "source": [
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=SVC(kernel='rbf',gamma = 10,C = 10, random_state=42)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BJ3aiISAnxhh",
        "outputId": "f9cfba70-ff90-4f0f-9ca6-7eedaf89114e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6478758169934641\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.65      0.69      0.67       604\n",
            "           1       0.80      0.88      0.84       809\n",
            "           2       0.31      0.41      0.35       173\n",
            "           3       0.62      0.48      0.54       219\n",
            "           4       0.79      0.59      0.68       234\n",
            "           5       0.34      0.26      0.30       148\n",
            "           6       0.28      0.27      0.27       118\n",
            "           7       0.65      0.49      0.56       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.55      0.51      0.53      2448\n",
            "weighted avg       0.65      0.65      0.64      2448\n",
            "\n",
            "F1 Score : 0.6968954248366013\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.69      0.82      0.75       604\n",
            "           1       0.87      0.89      0.88       809\n",
            "           2       0.40      0.37      0.38       174\n",
            "           3       0.59      0.54      0.56       218\n",
            "           4       0.74      0.72      0.73       234\n",
            "           5       0.41      0.31      0.35       148\n",
            "           6       0.33      0.24      0.27       118\n",
            "           7       0.61      0.50      0.55       143\n",
            "\n",
            "    accuracy                           0.70      2448\n",
            "   macro avg       0.58      0.55      0.56      2448\n",
            "weighted avg       0.68      0.70      0.69      2448\n",
            "\n",
            "F1 Score : 0.7246732026143791\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.74      0.82      0.78       604\n",
            "           1       0.89      0.88      0.88       808\n",
            "           2       0.42      0.30      0.35       174\n",
            "           3       0.63      0.69      0.66       218\n",
            "           4       0.74      0.77      0.76       234\n",
            "           5       0.42      0.38      0.40       149\n",
            "           6       0.43      0.36      0.39       118\n",
            "           7       0.62      0.59      0.60       143\n",
            "\n",
            "    accuracy                           0.72      2448\n",
            "   macro avg       0.61      0.60      0.60      2448\n",
            "weighted avg       0.71      0.72      0.72      2448\n",
            "\n",
            "F1 Score : 0.6548202614379085\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.74      0.65      0.69       604\n",
            "           1       0.86      0.82      0.84       808\n",
            "           2       0.25      0.20      0.22       173\n",
            "           3       0.51      0.55      0.53       219\n",
            "           4       0.61      0.85      0.71       233\n",
            "           5       0.35      0.45      0.39       149\n",
            "           6       0.38      0.42      0.40       119\n",
            "           7       0.64      0.52      0.57       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.54      0.56      0.55      2448\n",
            "weighted avg       0.66      0.65      0.66      2448\n",
            "\n",
            "F1 Score : 0.6405228758169934\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.49      0.60       604\n",
            "           1       0.83      0.88      0.85       808\n",
            "           2       0.36      0.31      0.34       173\n",
            "           3       0.46      0.67      0.55       219\n",
            "           4       0.79      0.67      0.72       233\n",
            "           5       0.33      0.47      0.39       149\n",
            "           6       0.38      0.28      0.32       118\n",
            "           7       0.43      0.74      0.55       144\n",
            "\n",
            "    accuracy                           0.64      2448\n",
            "   macro avg       0.54      0.56      0.54      2448\n",
            "weighted avg       0.67      0.64      0.64      2448\n",
            "\n",
            "F1 MEAN Score : 0.6729575163398693\n"
          ]
        }
      ],
      "source": [
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=LogisticRegression(solver='saga', n_jobs=1, C=1e5)\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eoJPzbN6pO36",
        "outputId": "c05ce4f4-bc64-4645-a3e5-c85b821696bf"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.5853758169934641\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.59      0.66      0.62       604\n",
            "           1       0.57      0.96      0.71       809\n",
            "           2       0.06      0.01      0.02       173\n",
            "           3       0.64      0.57      0.60       219\n",
            "           4       0.81      0.43      0.56       234\n",
            "           5       0.38      0.04      0.07       148\n",
            "           6       0.20      0.01      0.02       118\n",
            "           7       0.82      0.19      0.31       143\n",
            "\n",
            "    accuracy                           0.59      2448\n",
            "   macro avg       0.51      0.36      0.36      2448\n",
            "weighted avg       0.55      0.59      0.52      2448\n",
            "\n",
            "F1 Score : 0.6348039215686274\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.64      0.83      0.73       604\n",
            "           1       0.63      0.93      0.75       809\n",
            "           2       0.31      0.02      0.04       174\n",
            "           3       0.56      0.56      0.56       218\n",
            "           4       0.77      0.59      0.67       234\n",
            "           5       0.33      0.04      0.07       148\n",
            "           6       0.50      0.03      0.05       118\n",
            "           7       0.67      0.15      0.25       143\n",
            "\n",
            "    accuracy                           0.63      2448\n",
            "   macro avg       0.55      0.40      0.39      2448\n",
            "weighted avg       0.60      0.63      0.57      2448\n",
            "\n",
            "F1 Score : 0.6499183006535948\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.66      0.83      0.74       604\n",
            "           1       0.65      0.94      0.77       808\n",
            "           2       0.60      0.03      0.07       174\n",
            "           3       0.56      0.68      0.62       218\n",
            "           4       0.74      0.54      0.62       234\n",
            "           5       0.52      0.08      0.14       149\n",
            "           6       0.83      0.04      0.08       118\n",
            "           7       0.69      0.22      0.33       143\n",
            "\n",
            "    accuracy                           0.65      2448\n",
            "   macro avg       0.66      0.42      0.42      2448\n",
            "weighted avg       0.65      0.65      0.59      2448\n",
            "\n",
            "F1 Score : 0.6058006535947712\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.67      0.64      0.65       604\n",
            "           1       0.60      0.92      0.73       808\n",
            "           2       0.25      0.03      0.05       173\n",
            "           3       0.48      0.61      0.54       219\n",
            "           4       0.64      0.65      0.64       233\n",
            "           5       0.55      0.12      0.20       149\n",
            "           6       0.67      0.05      0.09       119\n",
            "           7       0.73      0.26      0.38       143\n",
            "\n",
            "    accuracy                           0.61      2448\n",
            "   macro avg       0.57      0.41      0.41      2448\n",
            "weighted avg       0.59      0.61      0.55      2448\n",
            "\n",
            "F1 Score : 0.5955882352941176\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.68      0.58      0.63       604\n",
            "           1       0.57      0.95      0.71       808\n",
            "           2       0.36      0.06      0.10       173\n",
            "           3       0.46      0.61      0.52       219\n",
            "           4       0.83      0.51      0.63       233\n",
            "           5       0.48      0.07      0.12       149\n",
            "           6       0.75      0.05      0.10       118\n",
            "           7       0.68      0.44      0.53       144\n",
            "\n",
            "    accuracy                           0.60      2448\n",
            "   macro avg       0.60      0.41      0.42      2448\n",
            "weighted avg       0.60      0.60      0.55      2448\n",
            "\n",
            "F1 MEAN Score : 0.6142973856209151\n"
          ]
        }
      ],
      "source": [
        "f1 = []\n",
        "\n",
        "for train_index, test_index in kfold.split(X_train, Y_train):\n",
        "\n",
        "    x_train, x_test = X_train[train_index], X_train[test_index]\n",
        "    y_train, y_test = Y_train[train_index], Y_train[test_index]\n",
        "    #print(x_train.shape, x_test.shape, y_train.shape, y_test.shape)\n",
        "    clf=RandomForestClassifier()\n",
        "    clf.fit(x_train, y_train)\n",
        "    clfval = clf.predict(x_test)   # predicting test data\n",
        "    print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))\n",
        "    print(classification_report(y_test, clfval))\n",
        "    f1.append(f1_score(y_test, clfval, average='micro'))\n",
        "print('F1 MEAN Score : {}'.format(np.mean(f1)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sq8YFGXzp8_E",
        "outputId": "02f0fcc8-6ebf-4432-cd1a-4679df6fb501"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.6556372549019608\n"
          ]
        }
      ],
      "source": [
        "from sklearn.svm import SVC\n",
        "clf=SVC(kernel='rbf',gamma = 1,C = 100, random_state=42)\n",
        "clf.fit(x_train, y_train)    # training model on train data\n",
        "\n",
        "clfval = clf.predict(x_test)   # predicting test data\n",
        "print('F1 Score : {}'.format(f1_score(y_test, clfval, average='micro')))  # printing F1 score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "XKRExfDvrQf3"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zp9bm_lhtLEX"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CEifkOV1tQKt"
      },
      "source": [
        "[link text](https://)# fine tunning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IA8M33Imtc6r",
        "outputId": "18eecfe3-6acd-4387-f35b-e027a613bc5c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[?25l\r\u001b[K     |█▊                              | 10 kB 21.4 MB/s eta 0:00:01\r\u001b[K     |███▍                            | 20 kB 17.1 MB/s eta 0:00:01\r\u001b[K     |█████                           | 30 kB 22.8 MB/s eta 0:00:01\r\u001b[K     |██████▊                         | 40 kB 14.4 MB/s eta 0:00:01\r\u001b[K     |████████▌                       | 51 kB 13.3 MB/s eta 0:00:01\r\u001b[K     |██████████▏                     | 61 kB 15.4 MB/s eta 0:00:01\r\u001b[K     |███████████▉                    | 71 kB 14.8 MB/s eta 0:00:01\r\u001b[K     |█████████████▌                  | 81 kB 14.2 MB/s eta 0:00:01\r\u001b[K     |███████████████▎                | 92 kB 15.6 MB/s eta 0:00:01\r\u001b[K     |█████████████████               | 102 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |██████████████████▋             | 112 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |████████████████████▎           | 122 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |██████████████████████          | 133 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |███████████████████████▊        | 143 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████▍      | 153 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████     | 163 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████▊   | 174 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████████▌ | 184 kB 13.5 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 193 kB 13.5 MB/s \n",
            "\u001b[?25h"
          ]
        }
      ],
      "source": [
        "! pip install torch skorch transformers -q"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "HpDarZmtt5N9"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "from sklearn.datasets import fetch_20newsgroups\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.pipeline import Pipeline\n",
        "from skorch import NeuralNetClassifier\n",
        "from skorch.callbacks import LRScheduler, ProgressBar\n",
        "from skorch.hf import HuggingfacePretrainedTokenizer\n",
        "from torch import nn\n",
        "from torch.optim.lr_scheduler import LambdaLR\n",
        "from transformers import AutoModelForSequenceClassification, AutoModel\n",
        "from transformers import AutoTokenizer"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tlVNbt1UPmn1"
      },
      "source": [
        "TOKENIZER = \"Shushant/nepaliBERT\"\n",
        "\n",
        "PRETRAINED_MODEL = \"Shushant/nepaliBERT\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "utNJg1bpuHmo"
      },
      "outputs": [],
      "source": [
        "# Choose a tokenizer and BERT model that work together\n",
        "TOKENIZER = \"NepBERTa/NepBERTa\"\n",
        "PRETRAINED_MODEL = \"NepBERTa/NepBERTa\"\n",
        "\n",
        "# model hyper-parameters\n",
        "OPTMIZER = torch.optim.AdamW\n",
        "LR = 5e-5\n",
        "MAX_EPOCHS = 4\n",
        "CRITERION = nn.BCELoss\n",
        "BATCH_SIZE = 8\n",
        "\n",
        "# device\n",
        "DEVICE = 'cuda' if torch.cuda.is_available() else 'cpu'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9izTmBwoutY7"
      },
      "outputs": [],
      "source": [
        "X = dftr['text'].iloc[:12240]\n",
        "y = np.array(y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vmiOnvYcAQRM",
        "outputId": "0efdc99c-b791-4c03-ff2d-bb1822681220"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0])"
            ]
          },
          "execution_count": 12,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y[1231]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PlrXLOAuvCaB"
      },
      "outputs": [],
      "source": [
        "X_train, X_test, y_train, y_test, = train_test_split(X, y, stratify=None, random_state=0,)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D6SckpEawGI8",
        "outputId": "d4d27d37-0bde-4490-fdbe-41cdf0a84102"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "8297    लकडाउन भन्थे॥। मान्छेको भिडमा लक भइन्छ भनेर पो...\n",
              "1231    कोभिड नेपाल: आन्तरिक उडान खोल्ने निर्णय, कसरी ...\n",
              "Name: text, dtype: object"
            ]
          },
          "execution_count": 46,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X_train[:2]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-xRzxunz_g9K",
        "outputId": "06b6634e-0332-4d38-db2b-df4ab513e666"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([[0, 0, 0, 1, 1, 0, 0, 0],\n",
              "       [0, 0, 0, 0, 0, 0, 0, 0]])"
            ]
          },
          "execution_count": 47,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_train[:2]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VfQerrhywSma"
      },
      "outputs": [],
      "source": [
        "num_training_steps = MAX_EPOCHS * (len(X_train) // BATCH_SIZE + 1)\n",
        "\n",
        "def lr_schedule(current_step):\n",
        "    factor = float(num_training_steps - current_step) / float(max(1, num_training_steps))\n",
        "    assert factor > 0\n",
        "    return factor"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zfaYfi2d9RdR"
      },
      "outputs": [],
      "source": [
        "'''def forward(self, ids, mask, token_type_ids):\n",
        "        _, features = self.roberta(ids, attention_mask = mask, token_type_ids = token_type_ids, return_dict=False)\n",
        "        output_2 = self.l2(features)\n",
        "        output = self.fc(output_2)\n",
        "        return output'''"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yzvl5u0Y-78P"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZeDUevemwumm"
      },
      "outputs": [],
      "source": [
        "class BertModule(nn.Module):\n",
        "    def __init__(self, name, num_labels):\n",
        "        super().__init__()\n",
        "        self.name = name\n",
        "        self.num_labels = num_labels\n",
        "        self.from_tf = True          #for tensorflow model(tf_model.h5)\n",
        "\n",
        "        self.reset_weights()\n",
        "\n",
        "    def reset_weights(self):\n",
        "        self.bert = AutoModel.from_pretrained(self.name,from_tf=self.from_tf)\n",
        "        self.classifier=nn.Linear(self.bert.config.hidden_size, self.num_labels)\n",
        "\n",
        "    def forward(self, **kwargs):\n",
        "        pred = self.bert(**kwargs)\n",
        "        pred = self.classifier(pred.pooler_output)\n",
        "        return pred"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "bWZNU48UwzL8"
      },
      "outputs": [],
      "source": [
        "pipeline = Pipeline([\n",
        "    ('tokenizer', HuggingfacePretrainedTokenizer(TOKENIZER)),\n",
        "    ('net', NeuralNetClassifier(\n",
        "        BertModule,\n",
        "        module__name=PRETRAINED_MODEL,\n",
        "        train_split=None,\n",
        "        module__num_labels=8,\n",
        "        optimizer=OPTMIZER,\n",
        "        lr=LR,\n",
        "        max_epochs=MAX_EPOCHS,\n",
        "        criterion=CRITERION,\n",
        "        batch_size=BATCH_SIZE,\n",
        "        iterator_train__shuffle=True,\n",
        "        device=DEVICE,\n",
        "        callbacks=[\n",
        "            LRScheduler(LambdaLR, lr_lambda=lr_schedule, step_every='batch'),\n",
        "            ProgressBar(),\n",
        "        ],\n",
        "    )),\n",
        "])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 575
        },
        "id": "YA4pBaicab-A",
        "outputId": "09b6db1b-c6f7-4870-9fe0-fb7d33ade964"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Re-initializing module because the following parameters were re-set: name, num_labels.\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "All TF 2.0 model weights were used when initializing BertModel.\n",
            "\n",
            "All the weights of BertModel were initialized from the TF 2.0 model.\n",
            "If your task is similar to the task the model of the checkpoint was trained on, you can already use BertModel for predictions without further training.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Re-initializing criterion.\n",
            "Re-initializing optimizer.\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 0/1148 [00:00<?, ?it/s]\n",
            "                                        \u001b[A"
          ]
        },
        {
          "ename": "RuntimeError",
          "evalue": "ignored",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
            "\u001b[0;32m<timed eval>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    392\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_final_estimator\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34m\"passthrough\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    393\u001b[0m                 \u001b[0mfit_params_last_step\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfit_params_steps\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msteps\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 394\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_final_estimator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mXt\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params_last_step\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    395\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    396\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/classifier.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    139\u001b[0m         \u001b[0;31m# this is actually a pylint bug:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    140\u001b[0m         \u001b[0;31m# https://github.com/PyCQA/pylint/issues/1085\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 141\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mNeuralNetClassifier\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    142\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    143\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpredict_proba\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m   1228\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minitialize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1229\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1230\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpartial_fit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1231\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1232\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mpartial_fit\u001b[0;34m(self, X, y, classes, **fit_params)\u001b[0m\n\u001b[1;32m   1187\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnotify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'on_train_begin'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1188\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1189\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_loop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1190\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1191\u001b[0m             \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mfit_loop\u001b[0;34m(self, X, y, epochs, **fit_params)\u001b[0m\n\u001b[1;32m   1100\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1101\u001b[0m             self.run_single_epoch(iterator_train, training=True, prefix=\"train\",\n\u001b[0;32m-> 1102\u001b[0;31m                                   step_fn=self.train_step, **fit_params)\n\u001b[0m\u001b[1;32m   1103\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m             self.run_single_epoch(iterator_valid, training=False, prefix=\"valid\",\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mrun_single_epoch\u001b[0;34m(self, iterator, training, prefix, step_fn, **fit_params)\u001b[0m\n\u001b[1;32m   1135\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mbatch\u001b[0m \u001b[0;32min\u001b[0m \u001b[0miterator\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1136\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnotify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"on_batch_begin\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtraining\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtraining\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1137\u001b[0;31m             \u001b[0mstep\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstep_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1138\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhistory\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecord_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mprefix\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"_loss\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstep\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"loss\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1139\u001b[0m             batch_size = (get_len(batch[0]) if isinstance(batch, (tuple, list))\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mtrain_step\u001b[0;34m(self, batch, **fit_params)\u001b[0m\n\u001b[1;32m   1014\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mstep\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'loss'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1015\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1016\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step_optimizer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_fn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1017\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mstep_accumulator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1018\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36m_step_optimizer\u001b[0;34m(self, step_fn)\u001b[0m\n\u001b[1;32m    970\u001b[0m                 \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    971\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 972\u001b[0;31m                 \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_fn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    973\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    974\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mtrain_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/lr_scheduler.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     63\u001b[0m                 \u001b[0minstance\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step_count\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     64\u001b[0m                 \u001b[0mwrapped\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__get__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minstance\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 65\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mwrapped\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     66\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     67\u001b[0m             \u001b[0;31m# Note that the returned function here is no longer a bound method,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/optimizer.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    111\u001b[0m                 \u001b[0mprofile_name\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"Optimizer.step#{}.step\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__class__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    112\u001b[0m                 \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecord_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mprofile_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 113\u001b[0;31m                     \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    114\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mwrapper\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    115\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/autograd/grad_mode.py\u001b[0m in \u001b[0;36mdecorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0;32mdef\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     26\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclone\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 27\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     28\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mcast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mF\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/adamw.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, closure)\u001b[0m\n\u001b[1;32m    117\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mclosure\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    118\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0menable_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 119\u001b[0;31m                 \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    120\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    121\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mgroup\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparam_groups\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mstep_fn\u001b[0;34m()\u001b[0m\n\u001b[1;32m   1004\u001b[0m         \u001b[0;32mdef\u001b[0m \u001b[0mstep_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1005\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_zero_grad_optimizer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1006\u001b[0;31m             \u001b[0mstep\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_step_single\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1007\u001b[0m             \u001b[0mstep_accumulator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstore_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1008\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mtrain_step_single\u001b[0;34m(self, batch, **fit_params)\u001b[0m\n\u001b[1;32m    904\u001b[0m         \u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0myi\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0munpack_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    905\u001b[0m         \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minfer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 906\u001b[0;31m         \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0myi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtraining\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    907\u001b[0m         \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    908\u001b[0m         return {\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/classifier.py\u001b[0m in \u001b[0;36mget_loss\u001b[0;34m(self, y_pred, y_true, *args, **kwargs)\u001b[0m\n\u001b[1;32m    124\u001b[0m             \u001b[0meps\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfinfo\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meps\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    125\u001b[0m             \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlog\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0meps\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 126\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    127\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    128\u001b[0m     \u001b[0;31m# pylint: disable=signature-differs\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mget_loss\u001b[0;34m(self, y_pred, y_true, X, training)\u001b[0m\n\u001b[1;32m   1569\u001b[0m         \"\"\"\n\u001b[1;32m   1570\u001b[0m         \u001b[0my_true\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mto_tensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevice\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1571\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcriterion_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1572\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1573\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mget_dataset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1128\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1129\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1130\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1131\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1132\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/loss.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input, target)\u001b[0m\n\u001b[1;32m    611\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    612\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 613\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mF\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbinary_cross_entropy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreduction\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreduction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    614\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    615\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py\u001b[0m in \u001b[0;36mbinary_cross_entropy\u001b[0;34m(input, target, weight, size_average, reduce, reduction)\u001b[0m\n\u001b[1;32m   3081\u001b[0m         \u001b[0mweight\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexpand\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnew_size\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3082\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3083\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_C\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_nn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbinary_cross_entropy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreduction_enum\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3084\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3085\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mRuntimeError\u001b[0m: Found dtype Long but expected Float"
          ]
        }
      ],
      "source": [
        "%time pipeline.fit(X_train, y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "shIKFCBpxKPX"
      },
      "outputs": [],
      "source": [
        "torch.manual_seed(42)\n",
        "torch.cuda.manual_seed(42)\n",
        "torch.cuda.manual_seed_all(42)\n",
        "np.random.seed(42)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 491
        },
        "id": "AZ2BpdIhxPED",
        "outputId": "591739c1-0429-4791-879b-6387223a1564"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Some weights of the model checkpoint at Sakonii/deberta-base-nepali were not used when initializing DebertaModel: ['cls.predictions.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.decoder.weight', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.decoder.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.dense.bias']\n",
            "- This IS expected if you are initializing DebertaModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
            "- This IS NOT expected if you are initializing DebertaModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
            "\n",
            "  0%|          | 0/1148 [00:00<?, ?it/s]\u001b[A"
          ]
        },
        {
          "ename": "AttributeError",
          "evalue": "ignored",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<timed eval>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    392\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_final_estimator\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34m\"passthrough\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    393\u001b[0m                 \u001b[0mfit_params_last_step\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfit_params_steps\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msteps\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 394\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_final_estimator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mXt\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params_last_step\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    395\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    396\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/classifier.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    139\u001b[0m         \u001b[0;31m# this is actually a pylint bug:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    140\u001b[0m         \u001b[0;31m# https://github.com/PyCQA/pylint/issues/1085\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 141\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mNeuralNetClassifier\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    142\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    143\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mpredict_proba\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m   1228\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minitialize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1229\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1230\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpartial_fit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1231\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1232\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mpartial_fit\u001b[0;34m(self, X, y, classes, **fit_params)\u001b[0m\n\u001b[1;32m   1187\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnotify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'on_train_begin'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1188\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1189\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_loop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1190\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1191\u001b[0m             \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mfit_loop\u001b[0;34m(self, X, y, epochs, **fit_params)\u001b[0m\n\u001b[1;32m   1100\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1101\u001b[0m             self.run_single_epoch(iterator_train, training=True, prefix=\"train\",\n\u001b[0;32m-> 1102\u001b[0;31m                                   step_fn=self.train_step, **fit_params)\n\u001b[0m\u001b[1;32m   1103\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m             self.run_single_epoch(iterator_valid, training=False, prefix=\"valid\",\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mrun_single_epoch\u001b[0;34m(self, iterator, training, prefix, step_fn, **fit_params)\u001b[0m\n\u001b[1;32m   1135\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mbatch\u001b[0m \u001b[0;32min\u001b[0m \u001b[0miterator\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1136\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnotify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"on_batch_begin\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtraining\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtraining\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1137\u001b[0;31m             \u001b[0mstep\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstep_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1138\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhistory\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecord_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mprefix\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"_loss\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstep\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"loss\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1139\u001b[0m             batch_size = (get_len(batch[0]) if isinstance(batch, (tuple, list))\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mtrain_step\u001b[0;34m(self, batch, **fit_params)\u001b[0m\n\u001b[1;32m   1014\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mstep\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'loss'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1015\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1016\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step_optimizer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_fn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1017\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mstep_accumulator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1018\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36m_step_optimizer\u001b[0;34m(self, step_fn)\u001b[0m\n\u001b[1;32m    970\u001b[0m                 \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    971\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 972\u001b[0;31m                 \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_fn\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    973\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    974\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mtrain_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/lr_scheduler.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     63\u001b[0m                 \u001b[0minstance\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_step_count\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     64\u001b[0m                 \u001b[0mwrapped\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__get__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minstance\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 65\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mwrapped\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     66\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     67\u001b[0m             \u001b[0;31m# Note that the returned function here is no longer a bound method,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/optimizer.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    111\u001b[0m                 \u001b[0mprofile_name\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"Optimizer.step#{}.step\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__class__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    112\u001b[0m                 \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprofiler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecord_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mprofile_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 113\u001b[0;31m                     \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    114\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mwrapper\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    115\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/autograd/grad_mode.py\u001b[0m in \u001b[0;36mdecorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0;32mdef\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     26\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclone\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 27\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     28\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mcast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mF\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdecorate_context\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/optim/adamw.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, closure)\u001b[0m\n\u001b[1;32m    117\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mclosure\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    118\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0menable_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 119\u001b[0;31m                 \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclosure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    120\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    121\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mgroup\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparam_groups\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mstep_fn\u001b[0;34m()\u001b[0m\n\u001b[1;32m   1004\u001b[0m         \u001b[0;32mdef\u001b[0m \u001b[0mstep_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1005\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_zero_grad_optimizer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1006\u001b[0;31m             \u001b[0mstep\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_step_single\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1007\u001b[0m             \u001b[0mstep_accumulator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstore_step\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1008\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36mtrain_step_single\u001b[0;34m(self, batch, **fit_params)\u001b[0m\n\u001b[1;32m    903\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_set_training\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    904\u001b[0m         \u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0myi\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0munpack_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 905\u001b[0;31m         \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minfer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    906\u001b[0m         \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0myi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mXi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtraining\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    907\u001b[0m         \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/skorch/net.py\u001b[0m in \u001b[0;36minfer\u001b[0;34m(self, x, **fit_params)\u001b[0m\n\u001b[1;32m   1424\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mMapping\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1425\u001b[0m             \u001b[0mx_dict\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_merge_x_and_fit_params\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1426\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodule_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mx_dict\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1427\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodule_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1428\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1128\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1129\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1130\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1131\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1132\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<ipython-input-147-9e0c13322f9e>\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, **kwargs)\u001b[0m\n\u001b[1;32m     15\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m         \u001b[0mpred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbert\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mpred\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlogits\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m: 'BaseModelOutput' object has no attribute 'logits'"
          ]
        }
      ],
      "source": [
        "%time pipeline.fit(X_train, y_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7CIMeiwJxSBB",
        "outputId": "a4ae90a1-28ac-4fb8-b56b-00250c1a75ce"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 1min, sys: 72.7 ms, total: 1min\n",
            "Wall time: 1min\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "with torch.inference_mode():\n",
        "    y_pred = pipeline.predict(X_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "klFNuYHIznkQ",
        "outputId": "c1559db6-efcc-4e8a-85cf-98b4fee97594"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0.834640522875817"
            ]
          },
          "execution_count": 51,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "accuracy_score(y_test, y_pred)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iT83HG3Yzujb",
        "outputId": "767ea1e4-a296-4987-83cb-de7903bd96b7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.834640522875817\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.83      1.00      0.91      2554\n",
            "           1       0.00      0.00      0.00       506\n",
            "\n",
            "    accuracy                           0.83      3060\n",
            "   macro avg       0.42      0.50      0.45      3060\n",
            "weighted avg       0.70      0.83      0.76      3060\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import confusion_matrix, f1_score, classification_report\n",
        "print('F1 Score : {}'.format(f1_score(y_test, y_pred, average='micro')))\n",
        "print(classification_report(y_test, y_pred))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 609
        },
        "id": "Ko4rgtJe0CiU",
        "outputId": "1102d0d4-04d9-4c7c-fc38-d36cea3bce43"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f61d6f3c2d0>"
            ]
          },
          "execution_count": 65,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix, f1_score\n",
        "cf_matrix = confusion_matrix(y_test, y_pred)\n",
        "fig, ax = plt.subplots(figsize=(15,10))\n",
        "sns.heatmap(cf_matrix, linewidths=1, annot=True, ax=ax, fmt='g')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OCR4cW4yM4-X"
      },
      "outputs": [],
      "source": [
        "# deberta"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cJDmaDlonfge",
        "outputId": "8452adf7-2ba5-4383-f740-f39cf248c939"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "F1 Score : 0.8137254901960784\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.86      0.77      0.81       755\n",
            "           1       0.95      0.99      0.97      1010\n",
            "           2       0.55      0.66      0.60       217\n",
            "           3       0.72      0.67      0.70       273\n",
            "           4       0.82      0.87      0.85       292\n",
            "           5       0.53      0.56      0.55       186\n",
            "           6       0.64      0.49      0.55       148\n",
            "           7       0.74      0.84      0.79       179\n",
            "\n",
            "    accuracy                           0.81      3060\n",
            "   macro avg       0.73      0.73      0.73      3060\n",
            "weighted avg       0.82      0.81      0.81      3060\n",
            "\n"
          ]
        }
      ],
      "source": [
        "print('F1 Score : {}'.format(f1_score(y_test, y_pred, average='micro')))\n",
        "print(classification_report(y_test, y_pred))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 612
        },
        "id": "JRyKi4bangMS",
        "outputId": "91171cbe-8987-4962-afec-7f6bbb66ca72"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f7a9a78e810>"
            ]
          },
          "execution_count": 24,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1080x720 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix\n",
        "\n",
        "cf_matrix = confusion_matrix(y_test, y_pred)\n",
        "fig, ax = plt.subplots(figsize=(15,10))\n",
        "sns.heatmap(cf_matrix, linewidths=1, annot=True, ax=ax, fmt='g')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GS_YvaTGnvCt"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d_9EcWGk2cwK"
      },
      "source": [
        "# tensorflow model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "p029caIE2lnE"
      },
      "outputs": [],
      "source": [
        "! pip install transformers -q"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ulunKB682cOI",
        "outputId": "d53edfa8-1aa3-4c38-b1ca-56f04dff39e3"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "TF version: 2.9.2\n"
          ]
        }
      ],
      "source": [
        "import os\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "import tensorflow as tf\n",
        "print(f'TF version: {tf.__version__}')\n",
        "from tensorflow.keras import layers\n",
        "import transformers"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "pHjXHPDM2iT4"
      },
      "outputs": [],
      "source": [
        "def set_seed(seed=42):\n",
        "    np.random.seed(seed)\n",
        "    tf.random.set_seed(seed)\n",
        "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
        "#     os.environ['TF_DETERMINISTIC_OPS'] = '1'\n",
        "set_seed(42)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 861
        },
        "id": "WXRLitt22uFZ",
        "outputId": "bf83f234-521c-42b9-d09a-81906fd26bb5"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-0f307344-7c7d-4553-96b2-ab9e4fbc6e6e\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>text</th>\n",
              "      <th>covid_stats</th>\n",
              "      <th>vaccination</th>\n",
              "      <th>covid_politics</th>\n",
              "      <th>humour</th>\n",
              "      <th>lockdown</th>\n",
              "      <th>civic_views</th>\n",
              "      <th>life_during_pandemic</th>\n",
              "      <th>covid_waves_and_variants</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>चितवनमा ९३ हजार बढीले लगाए कोरोनाविरुद्धको खोप</td>\n",
              "      <td>0</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>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>जोरबिजोर भनेको गाडी संख्या धेरै भएर ट्राफिक जा...</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>३१ सय ८ जना संक्रमित थपिदा १५ सय ९५ जना डिस्चा...</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>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>कोरोनाको जोखिम बढ्दै: झापाको दमक फेरी १ हफ्ता ...</td>\n",
              "      <td>0</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>4</th>\n",
              "      <td>कोरोना खोप राज्यले निःशुल्क लगाइरहेकै छ र थप ल...</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>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0f307344-7c7d-4553-96b2-ab9e4fbc6e6e')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-0f307344-7c7d-4553-96b2-ab9e4fbc6e6e button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-0f307344-7c7d-4553-96b2-ab9e4fbc6e6e');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "                                                text  covid_stats  \\\n",
              "0     चितवनमा ९३ हजार बढीले लगाए कोरोनाविरुद्धको खोप            0   \n",
              "1  जोरबिजोर भनेको गाडी संख्या धेरै भएर ट्राफिक जा...            0   \n",
              "2  ३१ सय ८ जना संक्रमित थपिदा १५ सय ९५ जना डिस्चा...            1   \n",
              "3  कोरोनाको जोखिम बढ्दै: झापाको दमक फेरी १ हफ्ता ...            0   \n",
              "4  कोरोना खोप राज्यले निःशुल्क लगाइरहेकै छ र थप ल...            0   \n",
              "\n",
              "   vaccination  covid_politics  humour  lockdown  civic_views  \\\n",
              "0            1               0       0         0            0   \n",
              "1            0               0       0         0            1   \n",
              "2            0               0       0         0            0   \n",
              "3            0               0       0         1            0   \n",
              "4            1               0       0         0            1   \n",
              "\n",
              "   life_during_pandemic  covid_waves_and_variants  \n",
              "0                     0                         1  \n",
              "1                     0                         0  \n",
              "2                     0                         0  \n",
              "3                     0                         0  \n",
              "4                     0                         0  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "---------DataFrame Summary---------\n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 12241 entries, 0 to 12240\n",
            "Data columns (total 9 columns):\n",
            " #   Column                    Non-Null Count  Dtype \n",
            "---  ------                    --------------  ----- \n",
            " 0   text                      12241 non-null  object\n",
            " 1   covid_stats               12241 non-null  int64 \n",
            " 2   vaccination               12241 non-null  int64 \n",
            " 3   covid_politics            12241 non-null  int64 \n",
            " 4   humour                    12241 non-null  int64 \n",
            " 5   lockdown                  12241 non-null  int64 \n",
            " 6   civic_views               12241 non-null  int64 \n",
            " 7   life_during_pandemic      12241 non-null  int64 \n",
            " 8   covid_waves_and_variants  12241 non-null  int64 \n",
            "dtypes: int64(8), object(1)\n",
            "memory usage: 860.8+ KB\n"
          ]
        }
      ],
      "source": [
        "df = pd.read_csv('https://github.com/naamiinepal/covid-tweet-classification/blob/main/training/datasets/nepali_tweets_dataset_labelled_tweets_feb_23.csv?raw=true')\n",
        "display(df.head())\n",
        "print('\\n---------DataFrame Summary---------')\n",
        "df.info()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aVhgsjuD23ss",
        "outputId": "334eaec9-7167-4ff2-909b-b6a8d8bb6ba9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Training examples: 9792, validation examples: 2449\n"
          ]
        }
      ],
      "source": [
        "train_df, val_df = train_test_split(df, test_size=0.2, random_state=42)\n",
        "print(f'Training examples: {len(train_df)}, validation examples: {len(val_df)}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nuhTxllK29vM"
      },
      "outputs": [],
      "source": [
        "TARGET_COLS = ['covid_stats', 'vaccination', 'covid_politics', 'humour', 'lockdown', 'civic_views','life_during_pandemic','covid_waves_and_variants']\n",
        "\n",
        "MAX_LENGTH = 512\n",
        "BATCH_SIZE = 8\n",
        "BERT_MODEL = \"NepBERTa/NepBERTa\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GGel6-153aaW"
      },
      "outputs": [],
      "source": [
        "class BertDataGenerator(tf.keras.utils.Sequence):\n",
        "    def __init__(\n",
        "        self,\n",
        "        text,\n",
        "        labels,\n",
        "        batch_size=BATCH_SIZE,\n",
        "        shuffle=True,\n",
        "        include_targets=True,\n",
        "    ):\n",
        "        self.text = text\n",
        "        self.labels = labels\n",
        "        self.shuffle = shuffle\n",
        "        self.batch_size = batch_size\n",
        "        self.include_targets = include_targets\n",
        "        self.tokenizer = transformers.BertTokenizer.from_pretrained(\n",
        "            BERT_MODEL, do_lower_case=True\n",
        "        )\n",
        "        self.indexes = np.arange(len(self.text))\n",
        "        self.on_epoch_end()\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.text) // self.batch_size\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        indexes = self.indexes[idx * self.batch_size : (idx + 1) * self.batch_size]\n",
        "        batch_texts = self.text[indexes]\n",
        "\n",
        "        encoded = self.tokenizer.batch_encode_plus(\n",
        "            batch_texts.tolist(),\n",
        "            add_special_tokens=True,\n",
        "            max_length=MAX_LENGTH,\n",
        "            return_attention_mask=True,\n",
        "            return_token_type_ids=True,\n",
        "            return_tensors=\"tf\",\n",
        "            truncation=True,\n",
        "            padding='max_length'\n",
        "        )\n",
        "\n",
        "        input_ids = np.array(encoded[\"input_ids\"], dtype=\"int32\")\n",
        "        attention_masks = np.array(encoded[\"attention_mask\"], dtype=\"int32\")\n",
        "        token_type_ids = np.array(encoded[\"token_type_ids\"], dtype=\"int32\")\n",
        "\n",
        "        if self.include_targets:\n",
        "            labels = np.array(self.labels[indexes], dtype=\"float32\")\n",
        "            return [input_ids, attention_masks, token_type_ids], labels\n",
        "        else:\n",
        "            return [input_ids, attention_masks, token_type_ids]\n",
        "\n",
        "    def on_epoch_end(self):\n",
        "        if self.shuffle:\n",
        "            np.random.RandomState(42).shuffle(self.indexes)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QGAYLx2I39UM"
      },
      "outputs": [],
      "source": [
        "train_data = BertDataGenerator(\n",
        "    train_df[\"text\"].values.astype(\"str\"),\n",
        "    np.array(train_df[TARGET_COLS]),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    shuffle=True,\n",
        ")\n",
        "valid_data = BertDataGenerator(\n",
        "    val_df[\"text\"].values.astype(\"str\"),\n",
        "    np.array(val_df[TARGET_COLS]),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    shuffle=False,\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b4Vmb2bQ4D5n",
        "outputId": "93dc0cc6-1bd9-48f8-ebe9-1f128e73c188"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "input_ids:\n",
            "[[    0  3400  2421 ...     1     1     1]\n",
            " [    0  2221  2175 ...     1     1     1]\n",
            " [    0  6411  4101 ...     1     1     1]\n",
            " ...\n",
            " [    0  2175 25274 ...     1     1     1]\n",
            " [    0  4229  1010 ...     1     1     1]\n",
            " [    0  3687  1026 ...     1     1     1]] \n",
            " With shape (8, 512) and dtype int32\n",
            "\n",
            "attention_mask:\n",
            "[[1 1 1 ... 0 0 0]\n",
            " [1 1 1 ... 0 0 0]\n",
            " [1 1 1 ... 0 0 0]\n",
            " ...\n",
            " [1 1 1 ... 0 0 0]\n",
            " [1 1 1 ... 0 0 0]\n",
            " [1 1 1 ... 0 0 0]] \n",
            " With shape (8, 512) and dtype int32\n",
            "\n",
            "token_type_ids:\n",
            "[[0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " ...\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]] \n",
            " With shape (8, 512) and dtype int32\n",
            "\n",
            "Labels:\n",
            "[[1. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 1. 0. 0. 0.]\n",
            " [0. 0. 0. 1. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 1. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 1. 0.]\n",
            " [0. 0. 0. 1. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 1. 0. 0.]] \n",
            " With shape (8, 8) and dtype float32\n"
          ]
        }
      ],
      "source": [
        "inputs, labels = next(iter(train_data))\n",
        "print(f'input_ids:\\n{inputs[0]} \\n With shape {inputs[0].shape} and dtype {inputs[0].dtype}\\n')\n",
        "print(f'attention_mask:\\n{inputs[1]} \\n With shape {inputs[1].shape} and dtype {inputs[0].dtype}\\n')\n",
        "print(f'token_type_ids:\\n{inputs[2]} \\n With shape {inputs[2].shape} and dtype {inputs[0].dtype}\\n')\n",
        "print(f'Labels:\\n{labels} \\n With shape {labels.shape} and dtype {labels.dtype}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "T50actkc4XEe"
      },
      "outputs": [],
      "source": [
        "def get_model():\n",
        "    input_ids = tf.keras.layers.Input(\n",
        "        shape=(MAX_LENGTH,), dtype=tf.int32, name=\"input_ids\"\n",
        "    )\n",
        "\n",
        "    attention_masks = tf.keras.layers.Input(\n",
        "        shape=(MAX_LENGTH,), dtype=tf.int32, name=\"attention_masks\"\n",
        "    )\n",
        "\n",
        "    token_type_ids = tf.keras.layers.Input(\n",
        "        shape=(MAX_LENGTH,), dtype=tf.int32, name=\"token_type_ids\"\n",
        "    )\n",
        "\n",
        "    bert_model = transformers.TFBertModel.from_pretrained(BERT_MODEL)\n",
        "    bert_model.trainable = True\n",
        "\n",
        "    bert_output = bert_model.bert(\n",
        "        input_ids, attention_mask=attention_masks, token_type_ids=token_type_ids\n",
        "    )\n",
        "    cls_output = bert_output.last_hidden_state[:, 0, :]\n",
        "    output = layers.Dense(8)(cls_output)\n",
        "    model = tf.keras.Model(inputs=[input_ids, attention_masks, token_type_ids], outputs=output)\n",
        "\n",
        "    return model\n",
        "\n",
        "\n",
        "optimizer=tf.optimizers.Adam(learning_rate=5e-5)\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LHKt_ZDJCUx8",
        "outputId": "16b8f118-7fe1-46ee-93de-5a95a7b10074"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting tensorflow_addons\n",
            "  Downloading tensorflow_addons-0.18.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.1 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.1 MB 24.4 MB/s \n",
            "\u001b[?25hRequirement already satisfied: typeguard>=2.7 in /usr/local/lib/python3.7/dist-packages (from tensorflow_addons) (2.7.1)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from tensorflow_addons) (21.3)\n",
            "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->tensorflow_addons) (3.0.9)\n",
            "Installing collected packages: tensorflow-addons\n",
            "Successfully installed tensorflow-addons-0.18.0\n"
          ]
        }
      ],
      "source": [
        "pip install tensorflow_addons"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PcRNtod9BsLl"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "import tensorflow_addons as tfa\n",
        "def get_loss_metrics():\n",
        "    loss = tf.keras.losses.BinaryCrossentropy(from_logits=True)\n",
        "    metrics = [\n",
        "        tf.metrics.BinaryAccuracy(),\n",
        "        tfa.metrics.F1Score(num_classes=8, average=\"weighted\", threshold=0.5),\n",
        "        tf.keras.metrics.AUC(\n",
        "            2449,\n",
        "            curve=\"PR\",\n",
        "            multi_label=True,\n",
        "            num_labels=8,\n",
        "            from_logits=True,\n",
        "        ),\n",
        "    ]\n",
        "\n",
        "    return loss, metrics"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "49hCKrGm6WqN",
        "outputId": "677272d8-cbba-4c6c-81e5-b79f04c76bec"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Some layers from the model checkpoint at NepBERTa/NepBERTa were not used when initializing TFBertModel: ['mlm___cls']\n",
            "- This IS expected if you are initializing TFBertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
            "- This IS NOT expected if you are initializing TFBertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
            "Some layers of TFBertModel were not initialized from the model checkpoint at NepBERTa/NepBERTa and are newly initialized: ['bert/pooler/dense/bias:0', 'bert/pooler/dense/kernel:0']\n",
            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_ids (InputLayer)         [(None, 512)]        0           []                               \n",
            "                                                                                                  \n",
            " attention_masks (InputLayer)   [(None, 512)]        0           []                               \n",
            "                                                                                                  \n",
            " token_type_ids (InputLayer)    [(None, 512)]        0           []                               \n",
            "                                                                                                  \n",
            " bert (TFBertMainLayer)         TFBaseModelOutputWi  109482240   ['input_ids[0][0]',              \n",
            "                                thPoolingAndCrossAt               'attention_masks[0][0]',        \n",
            "                                tentions(last_hidde               'token_type_ids[0][0]']         \n",
            "                                n_state=(None, 512,                                               \n",
            "                                 768),                                                            \n",
            "                                 pooler_output=(Non                                               \n",
            "                                e, 768),                                                          \n",
            "                                 past_key_values=No                                               \n",
            "                                ne, hidden_states=N                                               \n",
            "                                one, attentions=Non                                               \n",
            "                                e, cross_attentions                                               \n",
            "                                =None)                                                            \n",
            "                                                                                                  \n",
            " tf.__operators__.getitem (Slic  (None, 768)         0           ['bert[0][0]']                   \n",
            " ingOpLambda)                                                                                     \n",
            "                                                                                                  \n",
            " dense (Dense)                  (None, 8)            6152        ['tf.__operators__.getitem[0][0]'\n",
            "                                                                 ]                                \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 109,488,392\n",
            "Trainable params: 109,488,392\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "tf.keras.backend.clear_session()\n",
        "model = get_model()\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "811sszb3DR68"
      },
      "outputs": [],
      "source": [
        "loss, metrics = get_loss_metrics()\n",
        "model.compile(optimizer=optimizer, loss=loss, metrics=metrics)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RnpqEzBG6jjJ",
        "outputId": "171bbea1-3309-4d00-eae7-3aeb92f7b1ab"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Gradients do not exist for variables ['tf_bert_model/bert/pooler/dense/kernel:0', 'tf_bert_model/bert/pooler/dense/bias:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss`argument?\n",
            "WARNING:tensorflow:Gradients do not exist for variables ['tf_bert_model/bert/pooler/dense/kernel:0', 'tf_bert_model/bert/pooler/dense/bias:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss`argument?\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1224/1224 [==============================] - 1299s 1s/step - loss: 0.2240 - binary_accuracy: 0.9041 - f1_score: 0.6557 - auc: 0.7494 - val_loss: 0.2109 - val_binary_accuracy: 0.9114 - val_f1_score: 0.6737 - val_auc: 0.8044\n",
            "Epoch 2/5\n",
            "1224/1224 [==============================] - 1277s 1s/step - loss: 0.1700 - binary_accuracy: 0.9272 - f1_score: 0.7646 - auc: 0.8348 - val_loss: 0.1922 - val_binary_accuracy: 0.9201 - val_f1_score: 0.7248 - val_auc: 0.8367\n",
            "Epoch 3/5\n",
            "1224/1224 [==============================] - 1276s 1s/step - loss: 0.1531 - binary_accuracy: 0.9363 - f1_score: 0.8033 - auc: 0.8692 - val_loss: 0.1994 - val_binary_accuracy: 0.9152 - val_f1_score: 0.7088 - val_auc: 0.8284\n",
            "Epoch 4/5\n",
            "1224/1224 [==============================] - 1276s 1s/step - loss: 0.1268 - binary_accuracy: 0.9476 - f1_score: 0.8442 - auc: 0.9071 - val_loss: 0.1939 - val_binary_accuracy: 0.9287 - val_f1_score: 0.7901 - val_auc: 0.8386\n",
            "Epoch 5/5\n",
            " 275/1224 [=====>........................] - ETA: 15:09 - loss: 0.0909 - binary_accuracy: 0.9665 - f1_score: 0.9049 - auc: 0.9507"
          ]
        }
      ],
      "source": [
        "model.fit(train_data, validation_data=valid_data, epochs=5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "68tsCjjA7xhV",
        "outputId": "3987559e-7b1b-4f28-c001-b8dfe13667b7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "306/306 [==============================] - 111s 340ms/step\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "array([[-2.1905882 , -3.2394354 , -3.13258   , -2.1132884 , -2.4265406 ,\n",
              "        -0.6075255 , -1.2629524 , -1.5562562 ],\n",
              "       [-3.583075  ,  2.5596383 ,  0.3140234 , -5.727357  , -5.440543  ,\n",
              "        -5.8377833 , -5.9488435 , -3.1575963 ],\n",
              "       [-5.6693244 ,  2.5711992 ,  0.31066525, -4.5653663 , -4.1973977 ,\n",
              "        -2.9523754 , -4.2711267 , -4.5014157 ],\n",
              "       [-6.0214763 ,  2.1736732 , -0.54616076, -3.678521  , -3.931436  ,\n",
              "        -5.0702376 , -2.831023  , -2.5294201 ],\n",
              "       [-5.866425  ,  1.5135366 , -1.4845128 , -1.75077   , -2.6480722 ,\n",
              "        -0.91907746, -2.4754624 , -2.8797846 ]], dtype=float32)"
            ]
          },
          "execution_count": 63,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "pred = model.predict(valid_data)\n",
        "pred[:5]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z6_qTOKWGJhN",
        "outputId": "63cd1a07-d362-4c2d-c6ea-533939260043"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([[1, 0, 0, ..., 1, 0, 0],\n",
              "       [0, 1, 1, ..., 0, 0, 0],\n",
              "       [0, 1, 1, ..., 0, 0, 0],\n",
              "       ...,\n",
              "       [0, 1, 1, ..., 0, 0, 0],\n",
              "       [0, 1, 0, ..., 0, 0, 0],\n",
              "       [0, 0, 0, ..., 1, 0, 0]])"
            ]
          },
          "execution_count": 65,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "valid_data.labels"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7uM0EOfHHvYb",
        "outputId": "6a7b1ac7-4ce8-4290-f9b8-99e9122a9ed2"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "PreTrainedTokenizer(name_or_path='NepBERTa/NepBERTa', vocab_size=30523, model_max_len=1000000000000000019884624838656, is_fast=False, padding_side='right', truncation_side='right', special_tokens={'unk_token': '[UNK]', 'sep_token': '[SEP]', 'pad_token': '[PAD]', 'cls_token': '[CLS]', 'mask_token': '[MASK]'})"
            ]
          },
          "execution_count": 68,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 627
        },
        "id": "eBfa_mbEGvr6",
        "outputId": "8a7a1a36-bb0c-4705-c3f5-f245b2d15f7d"
      },
      "outputs": [
        {
          "ename": "ValueError",
          "evalue": "ignored",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-67-0e073d4760b0>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalid_data\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtext\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/keras/utils/traceback_utils.py\u001b[0m in \u001b[0;36merror_handler\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     65\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# pylint: disable=broad-except\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     66\u001b[0m       \u001b[0mfiltered_tb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_process_traceback_frames\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__traceback__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 67\u001b[0;31m       \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwith_traceback\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfiltered_tb\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     68\u001b[0m     \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     69\u001b[0m       \u001b[0;32mdel\u001b[0m \u001b[0mfiltered_tb\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mtf__predict_function\u001b[0;34m(iterator)\u001b[0m\n\u001b[1;32m     13\u001b[0m                 \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m                     \u001b[0mdo_return\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 15\u001b[0;31m                     \u001b[0mretval_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mld\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep_function\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mld\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mag__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mld\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     16\u001b[0m                 \u001b[0;32mexcept\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m                     \u001b[0mdo_return\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mValueError\u001b[0m: in user code:\n\n    File \"/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\", line 1845, in predict_function  *\n        return step_function(self, iterator)\n    File \"/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\", line 1834, in step_function  **\n        outputs = model.distribute_strategy.run(run_step, args=(data,))\n    File \"/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\", line 1823, in run_step  **\n        outputs = model.predict_step(data)\n    File \"/usr/local/lib/python3.7/dist-packages/keras/engine/training.py\", line 1791, in predict_step\n        return self(x, training=False)\n    File \"/usr/local/lib/python3.7/dist-packages/keras/utils/traceback_utils.py\", line 67, in error_handler\n        raise e.with_traceback(filtered_tb) from None\n    File \"/usr/local/lib/python3.7/dist-packages/keras/engine/input_spec.py\", line 200, in assert_input_compatibility\n        raise ValueError(f'Layer \"{layer_name}\" expects {len(input_spec)} input(s),'\n\n    ValueError: Layer \"model\" expects 3 input(s), but it received 1 input tensors. Inputs received: [<tf.Tensor 'IteratorGetNext:0' shape=(None,) dtype=string>]\n"
          ]
        }
      ],
      "source": [
        "model.predict(valid_data.text)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "irT-es5kHGHk"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GEDb6_BZOmIe"
      },
      "source": [
        "# multi label Bert type model finetuning"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AszFa0klO0C1",
        "outputId": "ecc08111-03b1-4825-bb7d-82348d048705"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[K     |████████████████████████████████| 5.5 MB 13.8 MB/s \n",
            "\u001b[K     |████████████████████████████████| 182 kB 75.6 MB/s \n",
            "\u001b[K     |████████████████████████████████| 7.6 MB 47.6 MB/s \n",
            "\u001b[?25h"
          ]
        }
      ],
      "source": [
        "!pip install transformers -q"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "0LdbRKQnOrOB"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import re\n",
        "import string\n",
        "import json\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn import metrics\n",
        "\n",
        "import transformers\n",
        "import torch\n",
        "from torch.utils.data import Dataset, DataLoader, RandomSampler, SequentialSampler\n",
        "from transformers import BertTokenizer, AutoTokenizer, BertModel, BertConfig, AutoModel, AdamW, get_linear_schedule_with_warmup, get_cosine_schedule_with_warmup\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "pd.set_option(\"display.max_columns\", None)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "Pswry9G_OweL",
        "outputId": "0c885dea-3633-45dd-e40c-9ba4266a3ff1"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "                                                Text  Category\n",
              "0  बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...  Politics\n",
              "1  राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...  Politics\n",
              "2  गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...  Politics\n",
              "3  गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...  Politics\n",
              "4  गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...  Politics"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d45d3fcc-64bc-48d5-bc2f-a7f0a8693652\" class=\"colab-df-container\">\n",
              "    <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>Text</th>\n",
              "      <th>Category</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...</td>\n",
              "      <td>Politics</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d45d3fcc-64bc-48d5-bc2f-a7f0a8693652')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-d45d3fcc-64bc-48d5-bc2f-a7f0a8693652 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-d45d3fcc-64bc-48d5-bc2f-a7f0a8693652');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-3260f962-591f-48d5-8df5-35ec39ec1754\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-3260f962-591f-48d5-8df5-35ec39ec1754')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-3260f962-591f-48d5-8df5-35ec39ec1754 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "df = pd.read_csv('/content/drive/MyDrive/merged_data.csv')\n",
        "df = df.dropna()\n",
        "df = df.reset_index(drop=True)\n",
        "display(df.head())"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create one-hot encoding for 'Category' with column names as unique values\n",
        "df_one_hot = pd.get_dummies(df['Category'])\n",
        "df_one_hot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 444
        },
        "id": "4_fNJOcndC7m",
        "outputId": "ae8bc913-7875-4173-c33f-9b1187120615"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Culture  Economy  EduTech  Entertainment  Health  Interview  \\\n",
              "0            0        0        0              0       0          0   \n",
              "1            0        0        0              0       0          0   \n",
              "2            0        0        0              0       0          0   \n",
              "3            0        0        0              0       0          0   \n",
              "4            0        0        0              0       0          0   \n",
              "...        ...      ...      ...            ...     ...        ...   \n",
              "15465        0        1        0              0       0          0   \n",
              "15466        0        1        0              0       0          0   \n",
              "15467        0        1        0              0       0          0   \n",
              "15468        0        1        0              0       0          0   \n",
              "15469        0        1        0              0       0          0   \n",
              "\n",
              "       Literature  Opinion  Politics  Sports  \n",
              "0               0        0         1       0  \n",
              "1               0        0         1       0  \n",
              "2               0        0         1       0  \n",
              "3               0        0         1       0  \n",
              "4               0        0         1       0  \n",
              "...           ...      ...       ...     ...  \n",
              "15465           0        0         0       0  \n",
              "15466           0        0         0       0  \n",
              "15467           0        0         0       0  \n",
              "15468           0        0         0       0  \n",
              "15469           0        0         0       0  \n",
              "\n",
              "[15470 rows x 10 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b9c1618c-0f08-4e29-a0b2-fb772eb69c71\" class=\"colab-df-container\">\n",
              "    <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>Culture</th>\n",
              "      <th>Economy</th>\n",
              "      <th>EduTech</th>\n",
              "      <th>Entertainment</th>\n",
              "      <th>Health</th>\n",
              "      <th>Interview</th>\n",
              "      <th>Literature</th>\n",
              "      <th>Opinion</th>\n",
              "      <th>Politics</th>\n",
              "      <th>Sports</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15465</th>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15466</th>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15467</th>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15468</th>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15469</th>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>15470 rows × 10 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b9c1618c-0f08-4e29-a0b2-fb772eb69c71')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-b9c1618c-0f08-4e29-a0b2-fb772eb69c71 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-b9c1618c-0f08-4e29-a0b2-fb772eb69c71');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-748d2f06-0426-4385-a936-4e709f374197\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-748d2f06-0426-4385-a936-4e709f374197')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-748d2f06-0426-4385-a936-4e709f374197 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Concatenate one-hot encoding with the original DataFrame\n",
        "df = pd.concat([df, df_one_hot], axis=1)"
      ],
      "metadata": {
        "id": "V37S2kv5eTLm"
      },
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 643
        },
        "id": "XAp5dLD-eVUy",
        "outputId": "948e4ea3-0535-4829-b5f3-620d2e3cf157"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                                Text  Category  Culture  \\\n",
              "0  बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...  Politics        0   \n",
              "1  राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...  Politics        0   \n",
              "2  गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...  Politics        0   \n",
              "3  गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...  Politics        0   \n",
              "4  गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...  Politics        0   \n",
              "\n",
              "   Economy  EduTech  Entertainment  Health  Interview  Literature  Opinion  \\\n",
              "0        0        0              0       0          0           0        0   \n",
              "1        0        0              0       0          0           0        0   \n",
              "2        0        0              0       0          0           0        0   \n",
              "3        0        0              0       0          0           0        0   \n",
              "4        0        0              0       0          0           0        0   \n",
              "\n",
              "   Politics  Sports  \n",
              "0         1       0  \n",
              "1         1       0  \n",
              "2         1       0  \n",
              "3         1       0  \n",
              "4         1       0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-637866f8-abb3-40ff-a312-104f513b3efd\" class=\"colab-df-container\">\n",
              "    <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>Text</th>\n",
              "      <th>Category</th>\n",
              "      <th>Culture</th>\n",
              "      <th>Economy</th>\n",
              "      <th>EduTech</th>\n",
              "      <th>Entertainment</th>\n",
              "      <th>Health</th>\n",
              "      <th>Interview</th>\n",
              "      <th>Literature</th>\n",
              "      <th>Opinion</th>\n",
              "      <th>Politics</th>\n",
              "      <th>Sports</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>बिहार मे होबय वाला नगर निगम \\nचुनाव 2022 के नव...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>राज्य मे भूमि सँ जुड़ल भ्रष्टाचार \\nकेर मामिला ...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>गांधीनगर | राष्ट्रीय एकता दिवस पर \\nसोम दिन गु...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>गुजरात केर मोरबी मे रवि दिन सांझ भेल दर्दनाक द...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>गुजरातक मोरबी\\nकेर उक्त झुलैत पुल \\nलगभग 150 स...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-637866f8-abb3-40ff-a312-104f513b3efd')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-637866f8-abb3-40ff-a312-104f513b3efd button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-637866f8-abb3-40ff-a312-104f513b3efd');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-d387a03b-516b-4b32-a3ef-a92f48e411b6\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-d387a03b-516b-4b32-a3ef-a92f48e411b6')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-d387a03b-516b-4b32-a3ef-a92f48e411b6 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rDIVLgyUO_g5",
        "outputId": "2201e38f-f506-404a-8cee-2ec91a71ae84"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training examples: 12376, validation examples: 3094\n"
          ]
        }
      ],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "df_train, df_dev = train_test_split(df, test_size=0.2, random_state=42)\n",
        "print(f'Training examples: {len(df_train)}, validation examples: {len(df_dev)}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "beXrYXW0PB7b",
        "outputId": "57b9b249-310c-43f0-870c-21d9f8d093f8"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       index                                               Text    Category  \\\n",
              "0      14374  पवन.- कार्टून मे उत्‍तराधिकारी सन कोनो चीज नहि...   Interview   \n",
              "1      14492  उत्तर – संगीत क विधिवत शिक्षा हम उस्ताद जौहर अ...   Interview   \n",
              "2       9144  3- बाल शिक्षासँ वंचित- परिवार में आर्थिक स्थित...  Literature   \n",
              "3       3268  एहिमे हिज्जेक गडबडी, शब्दक अशुद्धि, तकनीकी जे ...     Culture   \n",
              "4       2664  आ दोसर, जेना कि कतहु पढ़नहुँ छी जे पृथ्वी पर कत...     Culture   \n",
              "...      ...                                                ...         ...   \n",
              "12371   5191  राम जानकी विवाहपञ्चमी महोत्सव विशेषः अन्तिम वि...     Culture   \n",
              "12372  13418  \"दर्शन,– प्रवीण नारायण चौधरी,चलैत चलू – बढैत च...     Opinion   \n",
              "12373   5390  जिला विधिक सेवा प्राधिकरण के \\nतत्वावधान में म...     Culture   \n",
              "12374    860  वार्ड नं.२के मतगणना जारी छै । तथापि आएल मतगणना...    Politics   \n",
              "12375   7270  क्रिस्टियन रेसल रॉयलडो रोनाल्डो आरू अल्वारो मड...      Sports   \n",
              "\n",
              "       Culture  Economy  EduTech  Entertainment  Health  Interview  \\\n",
              "0            0        0        0              0       0          1   \n",
              "1            0        0        0              0       0          1   \n",
              "2            0        0        0              0       0          0   \n",
              "3            1        0        0              0       0          0   \n",
              "4            1        0        0              0       0          0   \n",
              "...        ...      ...      ...            ...     ...        ...   \n",
              "12371        1        0        0              0       0          0   \n",
              "12372        0        0        0              0       0          0   \n",
              "12373        1        0        0              0       0          0   \n",
              "12374        0        0        0              0       0          0   \n",
              "12375        0        0        0              0       0          0   \n",
              "\n",
              "       Literature  Opinion  Politics  Sports  \n",
              "0               0        0         0       0  \n",
              "1               0        0         0       0  \n",
              "2               1        0         0       0  \n",
              "3               0        0         0       0  \n",
              "4               0        0         0       0  \n",
              "...           ...      ...       ...     ...  \n",
              "12371           0        0         0       0  \n",
              "12372           0        1         0       0  \n",
              "12373           0        0         0       0  \n",
              "12374           0        0         1       0  \n",
              "12375           0        0         0       1  \n",
              "\n",
              "[12376 rows x 13 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1ff54c2e-54aa-4644-bc54-7d4525b3d82b\" class=\"colab-df-container\">\n",
              "    <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>index</th>\n",
              "      <th>Text</th>\n",
              "      <th>Category</th>\n",
              "      <th>Culture</th>\n",
              "      <th>Economy</th>\n",
              "      <th>EduTech</th>\n",
              "      <th>Entertainment</th>\n",
              "      <th>Health</th>\n",
              "      <th>Interview</th>\n",
              "      <th>Literature</th>\n",
              "      <th>Opinion</th>\n",
              "      <th>Politics</th>\n",
              "      <th>Sports</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>14374</td>\n",
              "      <td>पवन.- कार्टून मे उत्‍तराधिकारी सन कोनो चीज नहि...</td>\n",
              "      <td>Interview</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>14492</td>\n",
              "      <td>उत्तर – संगीत क विधिवत शिक्षा हम उस्ताद जौहर अ...</td>\n",
              "      <td>Interview</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>9144</td>\n",
              "      <td>3- बाल शिक्षासँ वंचित- परिवार में आर्थिक स्थित...</td>\n",
              "      <td>Literature</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3268</td>\n",
              "      <td>एहिमे हिज्जेक गडबडी, शब्दक अशुद्धि, तकनीकी जे ...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2664</td>\n",
              "      <td>आ दोसर, जेना कि कतहु पढ़नहुँ छी जे पृथ्वी पर कत...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12371</th>\n",
              "      <td>5191</td>\n",
              "      <td>राम जानकी विवाहपञ्चमी महोत्सव विशेषः अन्तिम वि...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12372</th>\n",
              "      <td>13418</td>\n",
              "      <td>\"दर्शन,– प्रवीण नारायण चौधरी,चलैत चलू – बढैत च...</td>\n",
              "      <td>Opinion</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>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12373</th>\n",
              "      <td>5390</td>\n",
              "      <td>जिला विधिक सेवा प्राधिकरण के \\nतत्वावधान में म...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12374</th>\n",
              "      <td>860</td>\n",
              "      <td>वार्ड नं.२के मतगणना जारी छै । तथापि आएल मतगणना...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12375</th>\n",
              "      <td>7270</td>\n",
              "      <td>क्रिस्टियन रेसल रॉयलडो रोनाल्डो आरू अल्वारो मड...</td>\n",
              "      <td>Sports</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>12376 rows × 13 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1ff54c2e-54aa-4644-bc54-7d4525b3d82b')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-1ff54c2e-54aa-4644-bc54-7d4525b3d82b button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-1ff54c2e-54aa-4644-bc54-7d4525b3d82b');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-7cdea122-bbb4-417d-a555-10ab40c591f6\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-7cdea122-bbb4-417d-a555-10ab40c591f6')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-7cdea122-bbb4-417d-a555-10ab40c591f6 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ],
      "source": [
        "df_train =  df_train.reset_index()\n",
        "df_train"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 913
        },
        "id": "_RnAcL-xPEyg",
        "outputId": "8c225fad-9d95-416a-d477-a618448b0c47"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      index                                               Text  Category  \\\n",
              "0      2519  ‘‘ देशमे अस्थिरता अशान्ति मचाबके काज नीयोजित र...  Politics   \n",
              "1      6025  । एजेंसी डेप्टो २०। रविवार क॑ बेन स्टोक्स न॑ द...    Sports   \n",
              "2      6366  । काठमांडू उपाध्यक्ष न्द बहादुर राज पसांग बैंग...    Sports   \n",
              "3      6962  । अक्टूबर के बाहर काठमांडू।नेपाली केरऽ कप्तान ...    Sports   \n",
              "4     13347  \"तुलसी कृपा रघुबंसमनि की लोह लै लौका तिरा॥\"\",\"...   Opinion   \n",
              "...     ...                                                ...       ...   \n",
              "3089  13860  \"बुद्धौ शरणमन्विच्छ कृपणाः फलहेतवः ॥४९॥\"\",एहि ...   Opinion   \n",
              "3090  12897  प्रश्न मैथिली भाषाक संरक्षणक अछि, प्रश्न नव रा...   EduTech   \n",
              "3091   5066  व्यक्तित्व परिचय एवं साक्षात्कार,– प्रवीण नारा...   Culture   \n",
              "3092   3975  \"मोहन भारद्वाज, मधुबनी। फरबरी १९, २०१७. मैथिली...   Culture   \n",
              "3093   3433  \"धरान, १६ दिसम्बर २०१८. मैथिली जिन्दाबाद!!\",\"क...   Culture   \n",
              "\n",
              "      Culture  Economy  EduTech  Entertainment  Health  Interview  Literature  \\\n",
              "0           0        0        0              0       0          0           0   \n",
              "1           0        0        0              0       0          0           0   \n",
              "2           0        0        0              0       0          0           0   \n",
              "3           0        0        0              0       0          0           0   \n",
              "4           0        0        0              0       0          0           0   \n",
              "...       ...      ...      ...            ...     ...        ...         ...   \n",
              "3089        0        0        0              0       0          0           0   \n",
              "3090        0        0        1              0       0          0           0   \n",
              "3091        1        0        0              0       0          0           0   \n",
              "3092        1        0        0              0       0          0           0   \n",
              "3093        1        0        0              0       0          0           0   \n",
              "\n",
              "      Opinion  Politics  Sports  \n",
              "0           0         1       0  \n",
              "1           0         0       1  \n",
              "2           0         0       1  \n",
              "3           0         0       1  \n",
              "4           1         0       0  \n",
              "...       ...       ...     ...  \n",
              "3089        1         0       0  \n",
              "3090        0         0       0  \n",
              "3091        0         0       0  \n",
              "3092        0         0       0  \n",
              "3093        0         0       0  \n",
              "\n",
              "[3094 rows x 13 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1ca157b3-0932-4531-a9f8-88caabee222c\" class=\"colab-df-container\">\n",
              "    <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>index</th>\n",
              "      <th>Text</th>\n",
              "      <th>Category</th>\n",
              "      <th>Culture</th>\n",
              "      <th>Economy</th>\n",
              "      <th>EduTech</th>\n",
              "      <th>Entertainment</th>\n",
              "      <th>Health</th>\n",
              "      <th>Interview</th>\n",
              "      <th>Literature</th>\n",
              "      <th>Opinion</th>\n",
              "      <th>Politics</th>\n",
              "      <th>Sports</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2519</td>\n",
              "      <td>‘‘ देशमे अस्थिरता अशान्ति मचाबके काज नीयोजित र...</td>\n",
              "      <td>Politics</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>6025</td>\n",
              "      <td>। एजेंसी डेप्टो २०। रविवार क॑ बेन स्टोक्स न॑ द...</td>\n",
              "      <td>Sports</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>6366</td>\n",
              "      <td>। काठमांडू उपाध्यक्ष न्द बहादुर राज पसांग बैंग...</td>\n",
              "      <td>Sports</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6962</td>\n",
              "      <td>। अक्टूबर के बाहर काठमांडू।नेपाली केरऽ कप्तान ...</td>\n",
              "      <td>Sports</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>13347</td>\n",
              "      <td>\"तुलसी कृपा रघुबंसमनि की लोह लै लौका तिरा॥\"\",\"...</td>\n",
              "      <td>Opinion</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>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3089</th>\n",
              "      <td>13860</td>\n",
              "      <td>\"बुद्धौ शरणमन्विच्छ कृपणाः फलहेतवः ॥४९॥\"\",एहि ...</td>\n",
              "      <td>Opinion</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>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3090</th>\n",
              "      <td>12897</td>\n",
              "      <td>प्रश्न मैथिली भाषाक संरक्षणक अछि, प्रश्न नव रा...</td>\n",
              "      <td>EduTech</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3091</th>\n",
              "      <td>5066</td>\n",
              "      <td>व्यक्तित्व परिचय एवं साक्षात्कार,– प्रवीण नारा...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3092</th>\n",
              "      <td>3975</td>\n",
              "      <td>\"मोहन भारद्वाज, मधुबनी। फरबरी १९, २०१७. मैथिली...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3093</th>\n",
              "      <td>3433</td>\n",
              "      <td>\"धरान, १६ दिसम्बर २०१८. मैथिली जिन्दाबाद!!\",\"क...</td>\n",
              "      <td>Culture</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3094 rows × 13 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1ca157b3-0932-4531-a9f8-88caabee222c')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-1ca157b3-0932-4531-a9f8-88caabee222c button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-1ca157b3-0932-4531-a9f8-88caabee222c');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-fb0bdbd0-507e-4b0c-9120-e2d7e399428e\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-fb0bdbd0-507e-4b0c-9120-e2d7e399428e')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-fb0bdbd0-507e-4b0c-9120-e2d7e399428e button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ],
      "source": [
        "df_dev =  df_dev.reset_index()\n",
        "df_dev"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yqw0hGtAPHPL",
        "outputId": "6adc1e96-e0ce-4289-ad28-2ff00f0c0196"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(12376, 13)\n",
            "(3094, 13)\n"
          ]
        }
      ],
      "source": [
        "print(df_train.shape)\n",
        "print(df_dev.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {
        "id": "E9wP-NzkPMn2"
      },
      "outputs": [],
      "source": [
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "id": "QVYt2itwPOiQ"
      },
      "outputs": [],
      "source": [
        "# Sections of config\n",
        "\n",
        "# Defining some key variables that will be used later on in the training\n",
        "MAX_LEN = 256\n",
        "TRAIN_BATCH_SIZE = 32\n",
        "VALID_BATCH_SIZE = 16\n",
        "EPOCHS = 4\n",
        "LEARNING_RATE = 5e-5\n",
        "tokenizer = AutoTokenizer.from_pretrained('NepBERTa/NepBERTa')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['Category'].unique()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L_lLrGv0Z5Aw",
        "outputId": "b8135e51-c0df-4b2e-9c0a-db4546e16392"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['Politics', 'Culture', 'Sports', 'Literature', 'Entertainment',\n",
              "       'Health', 'EduTech', 'Opinion', 'Interview', 'Economy'],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LhKyZsMMPTz4",
        "outputId": "44858eae-42ad-48ff-9f17-e7e972521aa2"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Politics',\n",
              " 'Culture',\n",
              " 'Sports',\n",
              " 'Literature',\n",
              " 'Entertainment',\n",
              " 'Health',\n",
              " 'EduTech',\n",
              " 'Opinion',\n",
              " 'Interview',\n",
              " 'Economy']"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ],
      "source": [
        "target_cols = ['Politics', 'Culture', 'Sports', 'Literature', 'Entertainment',\n",
        "       'Health', 'EduTech', 'Opinion', 'Interview', 'Economy']\n",
        "target_cols"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "id": "LhPmKFjkPV7J"
      },
      "outputs": [],
      "source": [
        "class BERTDataset(Dataset):\n",
        "    def __init__(self, df, tokenizer, max_len):\n",
        "        self.df = df\n",
        "        self.max_len = max_len\n",
        "        self.text = df.Text\n",
        "        self.tokenizer = tokenizer\n",
        "        self.targets = df[target_cols].values\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.df)\n",
        "\n",
        "    def __getitem__(self, index):\n",
        "        text = self.text[index]\n",
        "        inputs = self.tokenizer.encode_plus(\n",
        "            text,\n",
        "            truncation=True,\n",
        "            add_special_tokens=True,\n",
        "            max_length=self.max_len,\n",
        "            padding='max_length',\n",
        "            return_token_type_ids=True\n",
        "        )\n",
        "        ids = inputs['input_ids']\n",
        "        mask = inputs['attention_mask']\n",
        "        token_type_ids = inputs[\"token_type_ids\"]\n",
        "\n",
        "        return {\n",
        "            'ids': torch.tensor(ids, dtype=torch.long),\n",
        "            'mask': torch.tensor(mask, dtype=torch.long),\n",
        "            'token_type_ids': torch.tensor(token_type_ids, dtype=torch.long),\n",
        "            'targets': torch.tensor(self.targets[index], dtype=torch.float)\n",
        "        }"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 41,
      "metadata": {
        "id": "aHG6rpyZPYOc"
      },
      "outputs": [],
      "source": [
        "train_dataset = BERTDataset(df_train, tokenizer, MAX_LEN)\n",
        "valid_dataset = BERTDataset(df_dev, tokenizer, MAX_LEN)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 42,
      "metadata": {
        "id": "Z-2fqmrGPacn"
      },
      "outputs": [],
      "source": [
        "train_loader = DataLoader(train_dataset, batch_size=TRAIN_BATCH_SIZE,\n",
        "                          num_workers=4, shuffle=True, pin_memory=True)\n",
        "valid_loader = DataLoader(valid_dataset, batch_size=VALID_BATCH_SIZE,\n",
        "                          num_workers=4, shuffle=False, pin_memory=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 43,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1CoypLuyPeOD",
        "outputId": "26991222-ffa0-45fd-ce99-4ecc943158d2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "All TF 2.0 model weights were used when initializing BertModel.\n",
            "\n",
            "All the weights of BertModel were initialized from the TF 2.0 model.\n",
            "If your task is similar to the task the model of the checkpoint was trained on, you can already use BertModel for predictions without further training.\n"
          ]
        }
      ],
      "source": [
        "# Creating the customized model, by adding a drop out and a dense layer on top of distil bert to get the final output for the model.\n",
        "\n",
        "class BERTClass(torch.nn.Module):\n",
        "    def __init__(self):\n",
        "        super(BERTClass, self).__init__()\n",
        "        self.roberta = AutoModel.from_pretrained('NepBERTa/NepBERTa', from_tf = True)\n",
        "        self.l2 = torch.nn.Dropout(0.5)\n",
        "        self.fc = torch.nn.Linear(768,10)\n",
        "\n",
        "    def forward(self, ids, mask, token_type_ids):\n",
        "        _, features = self.roberta(ids, attention_mask = mask, token_type_ids = token_type_ids, return_dict=False)\n",
        "        output_2 = self.l2(features)\n",
        "        output = self.fc(output_2)\n",
        "        return output\n",
        "\n",
        "model = BERTClass()\n",
        "model.to(device);"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 44,
      "metadata": {
        "id": "dt8ex1foPkwh"
      },
      "outputs": [],
      "source": [
        "def loss_fn(outputs, targets):\n",
        "    return torch.nn.BCEWithLogitsLoss()(outputs, targets)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {
        "id": "xjr7wIG8Po-6"
      },
      "outputs": [],
      "source": [
        "optimizer = AdamW(params =  model.parameters(), lr=LEARNING_RATE, weight_decay=1e-6)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 46,
      "metadata": {
        "id": "cbxUdhgw8w43"
      },
      "outputs": [],
      "source": [
        "scheduler = get_cosine_schedule_with_warmup(\n",
        "                optimizer,\n",
        "                num_warmup_steps=0,\n",
        "                num_training_steps=len(train_loader) * EPOCHS\n",
        "            )"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 47,
      "metadata": {
        "id": "h4OHZ1IbPqlF"
      },
      "outputs": [],
      "source": [
        "def validation():\n",
        "    model.eval()\n",
        "    fin_targets=[]\n",
        "    fin_outputs=[]\n",
        "    with torch.no_grad():\n",
        "        for _, data in enumerate(valid_loader, 0):\n",
        "            ids = data['ids'].to(device, dtype = torch.long)\n",
        "            mask = data['mask'].to(device, dtype = torch.long)\n",
        "            token_type_ids = data['token_type_ids'].to(device, dtype = torch.long)\n",
        "            targets = data['targets'].to(device, dtype = torch.float)\n",
        "            outputs = model(ids, mask, token_type_ids)\n",
        "            fin_targets.extend(targets.cpu().detach().numpy().tolist())\n",
        "            fin_outputs.extend(torch.sigmoid(outputs).cpu().detach().numpy().tolist())\n",
        "    return fin_outputs, fin_targets"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 48,
      "metadata": {
        "id": "HcwI4uXePtjX"
      },
      "outputs": [],
      "source": [
        "def train(epoch):\n",
        "    model.train()\n",
        "    for _,data in enumerate(train_loader, 0):\n",
        "        ids = data['ids'].to(device, dtype = torch.long)\n",
        "        mask = data['mask'].to(device, dtype = torch.long)\n",
        "        token_type_ids = data['token_type_ids'].to(device, dtype = torch.long)\n",
        "        targets = data['targets'].to(device, dtype = torch.float)\n",
        "        outputs = model(ids, mask, token_type_ids)\n",
        "        loss = loss_fn(outputs, targets)\n",
        "        loss.backward()\n",
        "        torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm=1.0)\n",
        "        optimizer.step()\n",
        "        scheduler.step()\n",
        "        optimizer.zero_grad()\n",
        "    print(f'Epoch: {epoch}, Loss:  {loss.item()}')\n",
        "    outputs, targets = validation()\n",
        "    outputs = np.array(outputs) >= 0.5\n",
        "    accuracy = metrics.accuracy_score(targets, outputs)\n",
        "    f1_score_micro = metrics.f1_score(targets, outputs, average='micro')\n",
        "    f1_score_macro = metrics.f1_score(targets, outputs, average='macro')\n",
        "    print(f\"Accuracy Score = {accuracy}\")\n",
        "    print(f\"F1 Score (Micro) = {f1_score_micro}\")\n",
        "    print(f\"F1 Score (Macro) = {f1_score_macro}\")\n",
        "    print('\\n')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kRENHYzgPvzN",
        "outputId": "f1e3f3cd-f78b-425b-a751-1059dfeb0299"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0, Loss:  0.11394881457090378\n",
            "Accuracy Score = 0.7133160956690369\n",
            "F1 Score (Micro) = 0.7708697170799861\n",
            "F1 Score (Macro) = 0.6620882447650736\n",
            "\n",
            "\n",
            "Epoch: 1, Loss:  0.056467846035957336\n",
            "Accuracy Score = 0.7957336780866192\n",
            "F1 Score (Micro) = 0.8271004527922188\n",
            "F1 Score (Macro) = 0.7777480100592717\n",
            "\n",
            "\n",
            "Epoch: 2, Loss:  0.07918363064527512\n",
            "Accuracy Score = 0.8190045248868778\n",
            "F1 Score (Micro) = 0.8407622203811101\n",
            "F1 Score (Macro) = 0.8112309453618414\n",
            "\n",
            "\n",
            "Epoch: 3, Loss:  0.054994549602270126\n",
            "Accuracy Score = 0.8361344537815126\n",
            "F1 Score (Micro) = 0.853991769547325\n",
            "F1 Score (Macro) = 0.8326270924585992\n",
            "\n",
            "\n"
          ]
        }
      ],
      "source": [
        "for epoch in range(EPOCHS):\n",
        "    train(epoch)"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4fvdGfQLoxaG",
        "outputId": "37813162-0699-4aa2-b608-4dbbb15ba677"
      },
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 0, Loss:  0.11774052679538727\n",
            "Accuracy Score = 0.8387201034259858\n",
            "F1 Score (Micro) = 0.8543307086614174\n",
            "F1 Score (Macro) = 0.8331537892778929\n",
            "\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IiP-yv7RPz5C",
        "outputId": "b43fb649-b098-4dae-c8ae-e9428f9559c9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy Score = 0.8387201034259858\n",
            "F1 Score (Micro) = 0.8543307086614174\n",
            "F1 Score (Macro) = 0.8331537892778929\n",
            "F1 Score (Weighted) = 0.8537167986623573\n"
          ]
        }
      ],
      "source": [
        "outputs, targets = validation()\n",
        "outputs = np.array(outputs) >= 0.5\n",
        "accuracy = metrics.accuracy_score(targets, outputs)\n",
        "f1_score_micro = metrics.f1_score(targets, outputs, average='micro')\n",
        "f1_score_macro = metrics.f1_score(targets, outputs, average='macro')\n",
        "f1_score_weighted = metrics.f1_score(targets, outputs, average='weighted')\n",
        "print(f\"Accuracy Score = {accuracy}\")\n",
        "print(f\"F1 Score (Micro) = {f1_score_micro}\")\n",
        "print(f\"F1 Score (Macro) = {f1_score_macro}\")\n",
        "print(f\"F1 Score (Weighted) = {f1_score_weighted}\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 58,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C4E-5wGcQSYP",
        "outputId": "2e050b74-a5c4-4f56-b2eb-0a7e663b44dd"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9073170731707317 \n",
            " 0.8417721518987341 \n",
            " 0.9136690647482014 \n",
            " 0.8557213930348259 \n",
            " 0.8026607538802661 \n",
            " 0.7512690355329948 \n",
            " 0.641025641025641 \n",
            " 0.9005847953216374 \n",
            " 0.8383233532934132 \n",
            " 0.8791946308724832\n"
          ]
        }
      ],
      "source": [
        "print(metrics.f1_score(outputs[:,0],np.array(targets)[:,0]),'\\n',\n",
        "metrics.f1_score(outputs[:,1],np.array(targets)[:,1]),'\\n',\n",
        "metrics.f1_score(outputs[:,2],np.array(targets)[:,2]),'\\n',\n",
        "metrics.f1_score(outputs[:,3],np.array(targets)[:,3]),'\\n',\n",
        "metrics.f1_score(outputs[:,4],np.array(targets)[:,4]),'\\n',\n",
        "metrics.f1_score(outputs[:,5],np.array(targets)[:,5]),'\\n',\n",
        "metrics.f1_score(outputs[:,6],np.array(targets)[:,6]),'\\n',\n",
        "metrics.f1_score(outputs[:,7],np.array(targets)[:,7]),'\\n',\n",
        "metrics.f1_score(outputs[:,8],np.array(targets)[:,8]),'\\n',\n",
        "metrics.f1_score(outputs[:,9],np.array(targets)[:,9])\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 59,
      "metadata": {
        "id": "Jwc2h18NQWul"
      },
      "outputs": [],
      "source": [
        "torch.save(model.state_dict(), 'model.bin')"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.6"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "007304dd7fb74a1b9d3b390f13e73a03": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_821aa443392a4ac99d1adb7cc9138840",
            "max": 589,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_888b6d0488a649f7a4469b93f1bc7bc6",
            "value": 589
          }
        },
        "020a891f78e14676b691abd046356725": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0550c59aa57f40e094beae25cc5b0dc3": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "0766b7d42af943d0b4d37cab4bddf569": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "076956eb941c4cd6a402c1335aa3bba1": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0c31a369e5db4e0bbf689734f4dcd415": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0ddf0870774544bc9122761b7968c2ec": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "118705ff5aa84e59801ce43e13bf782f": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "14feea72286044c3ac19c79bdb4a5d15": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "18c49519429c41f195d703bed69a4443": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "1967dfefa2e24c76a680a450dfa66e6f": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "1a602627ce7742e7bf9af421ab365430": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "20b7ae08f2de4b86bda6c5f94054eb75": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "23531ca8f9e14e7dbc8f1ab522131142": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "2f56fa2517b2460a8d432ea866b657dd": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "32293f1a4a3e4b49b27f4be3fe5c3b1a": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_bc565a24c7fb4bb799eb45d251cc9a57",
            "placeholder": "​",
            "style": "IPY_MODEL_14feea72286044c3ac19c79bdb4a5d15",
            "value": " 413/413 [00:00&lt;00:00, 14.0kB/s]"
          }
        },
        "3367b5295e4b44d5a3aee0aaf5fbd73a": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "363c84ef86a34e6a93f4db2723a3c9f4": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_768426a2c2ee437389cdded02f3f0c48",
            "placeholder": "​",
            "style": "IPY_MODEL_44323c2a91184648aa0af06e91ab980d",
            "value": " 557M/557M [00:11&lt;00:00, 60.9MB/s]"
          }
        },
        "36620b6e478042989fe2ce175b627125": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "39a62a0780dc4911bd30d366ac81b6f3": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "3cdaa669bed14105a18bb904190b43a6": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_c9a5a153a6d8456fb867a2bbeee2368d",
            "placeholder": "​",
            "style": "IPY_MODEL_7d940db073774c888ab9a98316fbc3ae",
            "value": "Downloading: 100%"
          }
        },
        "3d7fc90750fc4fd28b2b6bef0ca6e601": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_0ddf0870774544bc9122761b7968c2ec",
            "placeholder": "​",
            "style": "IPY_MODEL_68e41c648aef493ab2290e9bd95ab244",
            "value": "Downloading: 100%"
          }
        },
        "4298e20b35474077a0f872ddbc6db987": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_20b7ae08f2de4b86bda6c5f94054eb75",
            "placeholder": "​",
            "style": "IPY_MODEL_879f999eba754a20861a37fcb275c7e0",
            "value": " 589/589 [00:00&lt;00:00, 6.83kB/s]"
          }
        },
        "44323c2a91184648aa0af06e91ab980d": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "4528dd10b894401e98dcaf0dd547bf13": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_d7892059dc6048b0bfa8e18998125727",
              "IPY_MODEL_8664160808b0488bba20727074611c9f",
              "IPY_MODEL_e9f2d971371a4c5cac7c9eab4ec30d89"
            ],
            "layout": "IPY_MODEL_511fdf809305494ca560dd7a84e8f904"
          }
        },
        "499f1d9965704b1a9601ca6701e886de": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_9541cd9ae94e469f8c0ed5ce66a0b01d",
            "placeholder": "​",
            "style": "IPY_MODEL_c4dbec2e026a4722bbf17bef4ad5741b",
            "value": "Downloading: 100%"
          }
        },
        "4b7ba118f59646e59cec9056ffcbcaa3": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "4d8556c2c15046f481701f853e13b126": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_8c3c9b20a3154dceb842acbf88160743",
              "IPY_MODEL_7dc6e325b5454fe294d3fe960fa8f2f0",
              "IPY_MODEL_363c84ef86a34e6a93f4db2723a3c9f4"
            ],
            "layout": "IPY_MODEL_d96a8fd5fb1540a0902e1fac49912602"
          }
        },
        "511fdf809305494ca560dd7a84e8f904": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "51bbf362fa7a49d0802a970b5443b01d": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_f183c27d35414a5facef4bed84aded9b",
            "placeholder": "​",
            "style": "IPY_MODEL_3367b5295e4b44d5a3aee0aaf5fbd73a",
            "value": " 2.13M/2.13M [00:00&lt;00:00, 2.72MB/s]"
          }
        },
        "5253b2271568488da290f23826fa8979": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "525d941c1e3541919bacb96f9adfdd91": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "54891ad613ed4f27bb3aa9be849132b1": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_4b7ba118f59646e59cec9056ffcbcaa3",
            "placeholder": "​",
            "style": "IPY_MODEL_525d941c1e3541919bacb96f9adfdd91",
            "value": "Downloading: 100%"
          }
        },
        "5d7a2ed7e1eb431ca509ede35d4e562f": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_0c31a369e5db4e0bbf689734f4dcd415",
            "placeholder": "​",
            "style": "IPY_MODEL_ec44446c5ce541e4a23d601aaa581a69",
            "value": "Downloading: 100%"
          }
        },
        "662f45e53d6d45f49c249cabc6f1dc91": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a79e620eb592460e986ea5c305ebeaec",
            "max": 2134747,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_39a62a0780dc4911bd30d366ac81b6f3",
            "value": 2134747
          }
        },
        "67c61a63e12e48dda06d76fd11a0d7e5": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "67fdea3f80f54b849854ec829e37d80b": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_118705ff5aa84e59801ce43e13bf782f",
            "placeholder": "​",
            "style": "IPY_MODEL_23531ca8f9e14e7dbc8f1ab522131142",
            "value": " 438M/438M [00:13&lt;00:00, 35.7MB/s]"
          }
        },
        "68e41c648aef493ab2290e9bd95ab244": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "6ebab858b53d466a96db54ced3956fd0": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "708178f346da4f1597853c0218a73e03": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "70d42517af9f49e38cf99b7c97b05239": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d3b1b8f83ae34e7d90ab862e23d2d5d9",
            "placeholder": "​",
            "style": "IPY_MODEL_9afec31fa2c542edbc191282ee03df73",
            "value": " 773/773 [00:00&lt;00:00, 10.1kB/s]"
          }
        },
        "714bdef5435d45ac8dbb7666c0e7232e": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_7ceed10a595343168a068bebf9746bca",
            "max": 413,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_67c61a63e12e48dda06d76fd11a0d7e5",
            "value": 413
          }
        },
        "748462ab70364e168c082c7573b52238": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_020a891f78e14676b691abd046356725",
            "max": 438141995,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_0550c59aa57f40e094beae25cc5b0dc3",
            "value": 438141995
          }
        },
        "768426a2c2ee437389cdded02f3f0c48": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "7a1044e4ca444ae990d163dff600a113": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "7ceed10a595343168a068bebf9746bca": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "7d940db073774c888ab9a98316fbc3ae": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "7dc6e325b5454fe294d3fe960fa8f2f0": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_708178f346da4f1597853c0218a73e03",
            "max": 557064301,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_c792e1ae15ce4fa29ccac9e75780bdd1",
            "value": 557064301
          }
        },
        "802d3e98e6b94a6881d3eb2f11ddb235": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "821aa443392a4ac99d1adb7cc9138840": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8445fd819a6a46fcbf75e99cbb18b4ce": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8664160808b0488bba20727074611c9f": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a8395f7eabf54ffc8329dfed96cc400a",
            "max": 528678,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_2f56fa2517b2460a8d432ea866b657dd",
            "value": 528678
          }
        },
        "879f999eba754a20861a37fcb275c7e0": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "888b6d0488a649f7a4469b93f1bc7bc6": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "8c3c9b20a3154dceb842acbf88160743": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_18c49519429c41f195d703bed69a4443",
            "placeholder": "​",
            "style": "IPY_MODEL_36620b6e478042989fe2ce175b627125",
            "value": "Downloading: 100%"
          }
        },
        "91b65e18dfae4ee08c23d8d70436f3d2": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_6ebab858b53d466a96db54ced3956fd0",
            "placeholder": "​",
            "style": "IPY_MODEL_1967dfefa2e24c76a680a450dfa66e6f",
            "value": " 239/239 [00:00&lt;00:00, 6.91kB/s]"
          }
        },
        "93fa5389c5e844d98eb849c08c035c1d": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "95400b69037c4c3b8ce7ada698d046ce": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9541cd9ae94e469f8c0ed5ce66a0b01d": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9afec31fa2c542edbc191282ee03df73": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "9dfcb1560dcd483ab5c000d6e43771f3": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_b5422f0f1b65405ea399b64dd602c2a4",
            "max": 239,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_93fa5389c5e844d98eb849c08c035c1d",
            "value": 239
          }
        },
        "a054cd75df6a40a1a8a23cb82c73e73d": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a2b590b036264886835d76ebae8de084": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_5d7a2ed7e1eb431ca509ede35d4e562f",
              "IPY_MODEL_714bdef5435d45ac8dbb7666c0e7232e",
              "IPY_MODEL_32293f1a4a3e4b49b27f4be3fe5c3b1a"
            ],
            "layout": "IPY_MODEL_0766b7d42af943d0b4d37cab4bddf569"
          }
        },
        "a360588788c34e59869144df00b459ec": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a42f03dc2a5849beb94666ff1648ff5b": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a79e620eb592460e986ea5c305ebeaec": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a8395f7eabf54ffc8329dfed96cc400a": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "ae16930d4b7e498e888c0b1a26ca9ecc": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "b1715cd48f6741088b3bf7b22f21030f": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_3d7fc90750fc4fd28b2b6bef0ca6e601",
              "IPY_MODEL_fcd6ab186c394b4fab091e2ec64cf7d2",
              "IPY_MODEL_70d42517af9f49e38cf99b7c97b05239"
            ],
            "layout": "IPY_MODEL_8445fd819a6a46fcbf75e99cbb18b4ce"
          }
        },
        "b5422f0f1b65405ea399b64dd602c2a4": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "bc565a24c7fb4bb799eb45d251cc9a57": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "c100b12a53de481ea1e5bce81995319c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_499f1d9965704b1a9601ca6701e886de",
              "IPY_MODEL_662f45e53d6d45f49c249cabc6f1dc91",
              "IPY_MODEL_51bbf362fa7a49d0802a970b5443b01d"
            ],
            "layout": "IPY_MODEL_1a602627ce7742e7bf9af421ab365430"
          }
        },
        "c4dbec2e026a4722bbf17bef4ad5741b": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "c792e1ae15ce4fa29ccac9e75780bdd1": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "c9a5a153a6d8456fb867a2bbeee2368d": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d0979c33cdc54526b5dd002b2c23bef6": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "d3b1b8f83ae34e7d90ab862e23d2d5d9": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d7892059dc6048b0bfa8e18998125727": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_802d3e98e6b94a6881d3eb2f11ddb235",
            "placeholder": "​",
            "style": "IPY_MODEL_a360588788c34e59869144df00b459ec",
            "value": "Downloading: 100%"
          }
        },
        "d7c45455a8e8445c929c8c83ab1f9fa1": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_f58360c8802b490fb23cb4a17a81911e",
            "placeholder": "​",
            "style": "IPY_MODEL_a42f03dc2a5849beb94666ff1648ff5b",
            "value": "Downloading: 100%"
          }
        },
        "d96a8fd5fb1540a0902e1fac49912602": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "dca828d0f4824d5a80125caf02b1bd18": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_3cdaa669bed14105a18bb904190b43a6",
              "IPY_MODEL_9dfcb1560dcd483ab5c000d6e43771f3",
              "IPY_MODEL_91b65e18dfae4ee08c23d8d70436f3d2"
            ],
            "layout": "IPY_MODEL_a054cd75df6a40a1a8a23cb82c73e73d"
          }
        },
        "dd083e9312f34cfe8712574bd16c2520": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_54891ad613ed4f27bb3aa9be849132b1",
              "IPY_MODEL_007304dd7fb74a1b9d3b390f13e73a03",
              "IPY_MODEL_4298e20b35474077a0f872ddbc6db987"
            ],
            "layout": "IPY_MODEL_076956eb941c4cd6a402c1335aa3bba1"
          }
        },
        "e9f2d971371a4c5cac7c9eab4ec30d89": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5253b2271568488da290f23826fa8979",
            "placeholder": "​",
            "style": "IPY_MODEL_ae16930d4b7e498e888c0b1a26ca9ecc",
            "value": " 529k/529k [00:00&lt;00:00, 1.41MB/s]"
          }
        },
        "ec44446c5ce541e4a23d601aaa581a69": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "f183c27d35414a5facef4bed84aded9b": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "f58360c8802b490fb23cb4a17a81911e": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "fcd6ab186c394b4fab091e2ec64cf7d2": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_95400b69037c4c3b8ce7ada698d046ce",
            "max": 773,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_d0979c33cdc54526b5dd002b2c23bef6",
            "value": 773
          }
        },
        "fe72571856854a17bf5df0b4ff3ff604": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_d7c45455a8e8445c929c8c83ab1f9fa1",
              "IPY_MODEL_748462ab70364e168c082c7573b52238",
              "IPY_MODEL_67fdea3f80f54b849854ec829e37d80b"
            ],
            "layout": "IPY_MODEL_7a1044e4ca444ae990d163dff600a113"
          }
        },
        "e96896658b654eed8b968404645d0ce2": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_18a4c50558434e0290dd7146a4fbfecf",
              "IPY_MODEL_84a98a576e0249a59849cc30c7224f20",
              "IPY_MODEL_e40536cd305045b0bb8aeecfd21cb956"
            ],
            "layout": "IPY_MODEL_1b15bace409c4f4089f40a856b6e1338"
          }
        },
        "18a4c50558434e0290dd7146a4fbfecf": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_74e25a277d244040ae97abc9dde27f68",
            "placeholder": "​",
            "style": "IPY_MODEL_eba06679f61f457b8de2817c7450ef24",
            "value": "config.json: 100%"
          }
        },
        "84a98a576e0249a59849cc30c7224f20": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_0b340f8a07e04471ac8ce7c17c61acba",
            "max": 652,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_4c87386df89a4157bf67851ad066bb94",
            "value": 652
          }
        },
        "e40536cd305045b0bb8aeecfd21cb956": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_f5ebefe5c7274db7b81dc329742a55e9",
            "placeholder": "​",
            "style": "IPY_MODEL_6fc5465c7f69419ebe26ac752133676f",
            "value": " 652/652 [00:00&lt;00:00, 34.7kB/s]"
          }
        },
        "1b15bace409c4f4089f40a856b6e1338": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "74e25a277d244040ae97abc9dde27f68": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "eba06679f61f457b8de2817c7450ef24": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "0b340f8a07e04471ac8ce7c17c61acba": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "4c87386df89a4157bf67851ad066bb94": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "f5ebefe5c7274db7b81dc329742a55e9": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6fc5465c7f69419ebe26ac752133676f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "4fcccd4ec6e54e608b823a93a59ac2b6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_a5ca839581e6414fba8dc778b5cf2cb9",
              "IPY_MODEL_a6dbe20598484fcd9f13675e697dfc7b",
              "IPY_MODEL_74f5680b28364d34a98ab4261563fc12"
            ],
            "layout": "IPY_MODEL_cbdb3c1dcc984f67a5856137830afe9c"
          }
        },
        "a5ca839581e6414fba8dc778b5cf2cb9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_4633fec68037492c92effc18055964e2",
            "placeholder": "​",
            "style": "IPY_MODEL_3b1fbd3a31f3422aa249db2d1c933a65",
            "value": "tf_model.h5: 100%"
          }
        },
        "a6dbe20598484fcd9f13675e697dfc7b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_bd2a23d5260345279fbf87befba2e5c8",
            "max": 533687680,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_6a2ed2b29770404a851a069f2a78c5da",
            "value": 533687680
          }
        },
        "74f5680b28364d34a98ab4261563fc12": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_bd8042817a254f35a6ff1c52c70cead2",
            "placeholder": "​",
            "style": "IPY_MODEL_f1a13f54dc4f48f8a613f522485755e1",
            "value": " 534M/534M [00:04&lt;00:00, 213MB/s]"
          }
        },
        "cbdb3c1dcc984f67a5856137830afe9c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "4633fec68037492c92effc18055964e2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "3b1fbd3a31f3422aa249db2d1c933a65": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "bd2a23d5260345279fbf87befba2e5c8": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6a2ed2b29770404a851a069f2a78c5da": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "bd8042817a254f35a6ff1c52c70cead2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "f1a13f54dc4f48f8a613f522485755e1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "c40ed648d38545e390bee2431c5a707a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_ee7cfc940c804296894abe79e31ecbe6",
              "IPY_MODEL_bc50f92e9d0d48bba21f16541fc37ed9",
              "IPY_MODEL_fe96466bdfcb41739954dc0270571406"
            ],
            "layout": "IPY_MODEL_69a754c07e454149aaa42ea63971f249"
          }
        },
        "ee7cfc940c804296894abe79e31ecbe6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_1ca2e994552f43558ee18ea4f832d2ec",
            "placeholder": "​",
            "style": "IPY_MODEL_b15d86eff02147bea51624bc82f94a7d",
            "value": "vocab.txt: 100%"
          }
        },
        "bc50f92e9d0d48bba21f16541fc37ed9": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_ccc737848a20419684a47c4dff68dc8a",
            "max": 546517,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_aad77fce445540cfbe295df700107b6a",
            "value": 546517
          }
        },
        "fe96466bdfcb41739954dc0270571406": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d7bce53523fd4cafb20182edcd405533",
            "placeholder": "​",
            "style": "IPY_MODEL_b6b326e3fe1148b29c5570b742e8b8c6",
            "value": " 547k/547k [00:00&lt;00:00, 16.3MB/s]"
          }
        },
        "69a754c07e454149aaa42ea63971f249": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "1ca2e994552f43558ee18ea4f832d2ec": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "b15d86eff02147bea51624bc82f94a7d": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "ccc737848a20419684a47c4dff68dc8a": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "aad77fce445540cfbe295df700107b6a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "d7bce53523fd4cafb20182edcd405533": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "b6b326e3fe1148b29c5570b742e8b8c6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}