SkimLit.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7VB4ERIhBn2h",
        "outputId": "1bb1d736-cb53-41ad-cf43-aa2607827afe"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "GPU 0: Tesla T4 (UUID: GPU-fca6208f-8307-7046-66c2-6f1b2c047c73)\n"
          ]
        }
      ],
      "source": [
        "!nvidia-smi -L"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!git clone https://github.com/Franck-Dernoncourt/pubmed-rct.git"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hgSpqLmlcNUj",
        "outputId": "5e71377a-61a1-47ec-b880-7a82e57d259c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "fatal: destination path 'pubmed-rct' already exists and is not an empty directory.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!ls pubmed-rct"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sn1Wr0D2s37Y",
        "outputId": "2baaa24e-a1af-430f-a063-9dec25b49b83"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "PubMed_200k_RCT\t\t\t\t       PubMed_20k_RCT_numbers_replaced_with_at_sign\n",
            "PubMed_200k_RCT_numbers_replaced_with_at_sign  README.md\n",
            "PubMed_20k_RCT\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# starting with the PubMed_20k_RCT\n",
        "!ls pubmed-rct/PubMed_20k_RCT_numbers_replaced_with_at_sign/"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vLErWudo_3Xu",
        "outputId": "ed853707-2c49-4ee2-db18-c77960e60446"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "dev.txt  test.txt  train.txt\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "data_dir = \"/content/pubmed-rct/PubMed_20k_RCT_numbers_replaced_with_at_sign/\""
      ],
      "metadata": {
        "id": "6zgT9T5aAYvP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# checking all the file names in the target directory\n",
        "import os\n",
        "filenames = [data_dir + filename for filename in os.listdir(data_dir)]\n",
        "filenames"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uSihu57CES5L",
        "outputId": "051ee173-0417-4534-9bd6-6d5d7645c6f8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['/content/pubmed-rct/PubMed_20k_RCT_numbers_replaced_with_at_sign/test.txt',\n",
              " '/content/pubmed-rct/PubMed_20k_RCT_numbers_replaced_with_at_sign/dev.txt',\n",
              " '/content/pubmed-rct/PubMed_20k_RCT_numbers_replaced_with_at_sign/train.txt']"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Preprocessing the data"
      ],
      "metadata": {
        "id": "oYdIFP8rE-Lb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_lines(filename):\n",
        "  \"\"\"\n",
        "  Reads filename (a text filename) and returns the lines of text as a list\n",
        "  \"\"\"\n",
        "  with open(filename, \"r\") as f:\n",
        "    return f.readlines()"
      ],
      "metadata": {
        "id": "vBBpmUScEjGp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_lines = get_lines(data_dir+\"train.txt\")\n",
        "train_lines[:27]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E75AMElZHnAN",
        "outputId": "e35087a2-f80a-44e3-d2fa-a2af6c8a12a1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['###24293578\\n',\n",
              " 'OBJECTIVE\\tTo investigate the efficacy of @ weeks of daily low-dose oral prednisolone in improving pain , mobility , and systemic low-grade inflammation in the short term and whether the effect would be sustained at @ weeks in older adults with moderate to severe knee osteoarthritis ( OA ) .\\n',\n",
              " 'METHODS\\tA total of @ patients with primary knee OA were randomized @:@ ; @ received @ mg/day of prednisolone and @ received placebo for @ weeks .\\n',\n",
              " 'METHODS\\tOutcome measures included pain reduction and improvement in function scores and systemic inflammation markers .\\n',\n",
              " 'METHODS\\tPain was assessed using the visual analog pain scale ( @-@ mm ) .\\n',\n",
              " 'METHODS\\tSecondary outcome measures included the Western Ontario and McMaster Universities Osteoarthritis Index scores , patient global assessment ( PGA ) of the severity of knee OA , and @-min walk distance ( @MWD ) .\\n',\n",
              " 'METHODS\\tSerum levels of interleukin @ ( IL-@ ) , IL-@ , tumor necrosis factor ( TNF ) - , and high-sensitivity C-reactive protein ( hsCRP ) were measured .\\n',\n",
              " 'RESULTS\\tThere was a clinically relevant reduction in the intervention group compared to the placebo group for knee pain , physical function , PGA , and @MWD at @ weeks .\\n',\n",
              " 'RESULTS\\tThe mean difference between treatment arms ( @ % CI ) was @ ( @-@ @ ) , p < @ ; @ ( @-@ @ ) , p < @ ; @ ( @-@ @ ) , p < @ ; and @ ( @-@ @ ) , p < @ , respectively .\\n',\n",
              " 'RESULTS\\tFurther , there was a clinically relevant reduction in the serum levels of IL-@ , IL-@ , TNF - , and hsCRP at @ weeks in the intervention group when compared to the placebo group .\\n',\n",
              " 'RESULTS\\tThese differences remained significant at @ weeks .\\n',\n",
              " 'RESULTS\\tThe Outcome Measures in Rheumatology Clinical Trials-Osteoarthritis Research Society International responder rate was @ % in the intervention group and @ % in the placebo group ( p < @ ) .\\n',\n",
              " 'CONCLUSIONS\\tLow-dose oral prednisolone had both a short-term and a longer sustained effect resulting in less knee pain , better physical function , and attenuation of systemic inflammation in older patients with knee OA ( ClinicalTrials.gov identifier NCT@ ) .\\n',\n",
              " '\\n',\n",
              " '###24854809\\n',\n",
              " 'BACKGROUND\\tEmotional eating is associated with overeating and the development of obesity .\\n',\n",
              " 'BACKGROUND\\tYet , empirical evidence for individual ( trait ) differences in emotional eating and cognitive mechanisms that contribute to eating during sad mood remain equivocal .\\n',\n",
              " 'OBJECTIVE\\tThe aim of this study was to test if attention bias for food moderates the effect of self-reported emotional eating during sad mood ( vs neutral mood ) on actual food intake .\\n',\n",
              " 'OBJECTIVE\\tIt was expected that emotional eating is predictive of elevated attention for food and higher food intake after an experimentally induced sad mood and that attentional maintenance on food predicts food intake during a sad versus a neutral mood .\\n',\n",
              " 'METHODS\\tParticipants ( N = @ ) were randomly assigned to one of the two experimental mood induction conditions ( sad/neutral ) .\\n',\n",
              " 'METHODS\\tAttentional biases for high caloric foods were measured by eye tracking during a visual probe task with pictorial food and neutral stimuli .\\n',\n",
              " 'METHODS\\tSelf-reported emotional eating was assessed with the Dutch Eating Behavior Questionnaire ( DEBQ ) and ad libitum food intake was tested by a disguised food offer .\\n',\n",
              " 'RESULTS\\tHierarchical multivariate regression modeling showed that self-reported emotional eating did not account for changes in attention allocation for food or food intake in either condition .\\n',\n",
              " 'RESULTS\\tYet , attention maintenance on food cues was significantly related to increased intake specifically in the neutral condition , but not in the sad mood condition .\\n',\n",
              " 'CONCLUSIONS\\tThe current findings show that self-reported emotional eating ( based on the DEBQ ) might not validly predict who overeats when sad , at least not in a laboratory setting with healthy women .\\n',\n",
              " 'CONCLUSIONS\\tResults further suggest that attention maintenance on food relates to eating motivation when in a neutral affective state , and might therefore be a cognitive mechanism contributing to increased food intake in general , but maybe not during sad mood .\\n',\n",
              " '\\n']"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_lines)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dsO-o_kEK_L7",
        "outputId": "15897dbb-5023-4658-a7cc-edffff4d198a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "210040"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# How we want our data to look\n",
        "```\n",
        "[\n",
        "  {'line_number': 0,\n",
        "    'target': 'BACKGROUND'\n",
        "    'text': \"Emotional eating is associated with overeating and the development of obesity .\\n\"\n",
        "    'total_lines': 11,\n",
        "    ...}\n",
        "]\n",
        "```"
      ],
      "metadata": {
        "id": "h9pqcrRBKhnx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def preprocess_text_with_line_numbers(filename):\n",
        "  \"\"\"Returns a list of dictionaries of abstract line data.\n",
        "\n",
        "  Takes in filename, reads its contents and sorts through each line,\n",
        "  extracting things like the target label, the text of the sentence,\n",
        "  how many sentences are in the current abstract and what sentence number\n",
        "  the target line is.\n",
        "\n",
        "  Args:\n",
        "      filename: a string of the target text file to read and extract line data\n",
        "      from.\n",
        "\n",
        "  Returns:\n",
        "      A list of dictionaries each containing a line from an abstract,\n",
        "      the lines label, the lines position in the abstract and the total number\n",
        "      of lines in the abstract where the line is from. For example:\n",
        "\n",
        "      [{\"target\": 'CONCLUSION',\n",
        "        \"text\": The study couldn't have gone better, turns out people are kinder than you think\",\n",
        "        \"line_number\": 8,\n",
        "        \"total_lines\": 8}]\n",
        "  \"\"\"\n",
        "  input_lines = get_lines(filename) # get all lines from filename\n",
        "  abstract_lines = \"\" # create an empty abstract\n",
        "  abstract_samples = [] # create an empty list of abstracts\n",
        "\n",
        "  # Loop through each line in target file\n",
        "  for line in input_lines:\n",
        "    if line.startswith(\"###\"): # check to see if line is an ID line\n",
        "      abstract_id = line\n",
        "      abstract_lines = \"\" # reset abstract string\n",
        "    elif line.isspace(): # check to see if line is a new line\n",
        "      abstract_line_split = abstract_lines.splitlines() # split abstract into separate lines\n",
        "\n",
        "      # Iterate through each line in abstract and count them at the same time\n",
        "      for abstract_line_number, abstract_line in enumerate(abstract_line_split):\n",
        "        line_data = {} # create empty dict to store data from line\n",
        "        target_text_split = abstract_line.split(\"\\t\") # split target label from text\n",
        "        line_data[\"target\"] = target_text_split[0] # get target label\n",
        "        line_data[\"text\"] = target_text_split[1].lower() # get target text and lower it\n",
        "        line_data[\"line_number\"] = abstract_line_number # what number line does the line appear in the abstract?\n",
        "        line_data[\"total_lines\"] = len(abstract_line_split) - 1 # how many total lines are in the abstract? (start from 0)\n",
        "        abstract_samples.append(line_data) # add line data to abstract samples list\n",
        "\n",
        "    else: # if the above conditions aren't fulfilled, the line contains a labelled sentence\n",
        "      abstract_lines += line\n",
        "\n",
        "  return abstract_samples"
      ],
      "metadata": {
        "id": "zJ1leUu9HoVt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "train_samples = preprocess_text_with_line_numbers(data_dir+\"train.txt\")\n",
        "val_samples = preprocess_text_with_line_numbers(data_dir+\"dev.txt\")\n",
        "test_samples = preprocess_text_with_line_numbers(data_dir+\"test.txt\")\n",
        "len(train_samples), len(val_samples), len(test_samples)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6noMwyqQSXPR",
        "outputId": "c77d3a23-acd9-434a-f7cc-07c29b5e74b5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CPU times: user 793 ms, sys: 135 ms, total: 928 ms\n",
            "Wall time: 1.09 s\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(180040, 30212, 30135)"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_samples[:14]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B1sBxZ93SfZq",
        "outputId": "1f2bd51f-b328-4363-c01b-0fb621334917"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[{'target': 'OBJECTIVE',\n",
              "  'text': 'to investigate the efficacy of @ weeks of daily low-dose oral prednisolone in improving pain , mobility , and systemic low-grade inflammation in the short term and whether the effect would be sustained at @ weeks in older adults with moderate to severe knee osteoarthritis ( oa ) .',\n",
              "  'line_number': 0,\n",
              "  'total_lines': 11},\n",
              " {'target': 'METHODS',\n",
              "  'text': 'a total of @ patients with primary knee oa were randomized @:@ ; @ received @ mg/day of prednisolone and @ received placebo for @ weeks .',\n",
              "  'line_number': 1,\n",
              "  'total_lines': 11},\n",
              " {'target': 'METHODS',\n",
              "  'text': 'outcome measures included pain reduction and improvement in function scores and systemic inflammation markers .',\n",
              "  'line_number': 2,\n",
              "  'total_lines': 11},\n",
              " {'target': 'METHODS',\n",
              "  'text': 'pain was assessed using the visual analog pain scale ( @-@ mm ) .',\n",
              "  'line_number': 3,\n",
              "  'total_lines': 11},\n",
              " {'target': 'METHODS',\n",
              "  'text': 'secondary outcome measures included the western ontario and mcmaster universities osteoarthritis index scores , patient global assessment ( pga ) of the severity of knee oa , and @-min walk distance ( @mwd ) .',\n",
              "  'line_number': 4,\n",
              "  'total_lines': 11},\n",
              " {'target': 'METHODS',\n",
              "  'text': 'serum levels of interleukin @ ( il-@ ) , il-@ , tumor necrosis factor ( tnf ) - , and high-sensitivity c-reactive protein ( hscrp ) were measured .',\n",
              "  'line_number': 5,\n",
              "  'total_lines': 11},\n",
              " {'target': 'RESULTS',\n",
              "  'text': 'there was a clinically relevant reduction in the intervention group compared to the placebo group for knee pain , physical function , pga , and @mwd at @ weeks .',\n",
              "  'line_number': 6,\n",
              "  'total_lines': 11},\n",
              " {'target': 'RESULTS',\n",
              "  'text': 'the mean difference between treatment arms ( @ % ci ) was @ ( @-@ @ ) , p < @ ; @ ( @-@ @ ) , p < @ ; @ ( @-@ @ ) , p < @ ; and @ ( @-@ @ ) , p < @ , respectively .',\n",
              "  'line_number': 7,\n",
              "  'total_lines': 11},\n",
              " {'target': 'RESULTS',\n",
              "  'text': 'further , there was a clinically relevant reduction in the serum levels of il-@ , il-@ , tnf - , and hscrp at @ weeks in the intervention group when compared to the placebo group .',\n",
              "  'line_number': 8,\n",
              "  'total_lines': 11},\n",
              " {'target': 'RESULTS',\n",
              "  'text': 'these differences remained significant at @ weeks .',\n",
              "  'line_number': 9,\n",
              "  'total_lines': 11},\n",
              " {'target': 'RESULTS',\n",
              "  'text': 'the outcome measures in rheumatology clinical trials-osteoarthritis research society international responder rate was @ % in the intervention group and @ % in the placebo group ( p < @ ) .',\n",
              "  'line_number': 10,\n",
              "  'total_lines': 11},\n",
              " {'target': 'CONCLUSIONS',\n",
              "  'text': 'low-dose oral prednisolone had both a short-term and a longer sustained effect resulting in less knee pain , better physical function , and attenuation of systemic inflammation in older patients with knee oa ( clinicaltrials.gov identifier nct@ ) .',\n",
              "  'line_number': 11,\n",
              "  'total_lines': 11},\n",
              " {'target': 'BACKGROUND',\n",
              "  'text': 'emotional eating is associated with overeating and the development of obesity .',\n",
              "  'line_number': 0,\n",
              "  'total_lines': 10},\n",
              " {'target': 'BACKGROUND',\n",
              "  'text': 'yet , empirical evidence for individual ( trait ) differences in emotional eating and cognitive mechanisms that contribute to eating during sad mood remain equivocal .',\n",
              "  'line_number': 1,\n",
              "  'total_lines': 10}]"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Visualising our data with pandas"
      ],
      "metadata": {
        "id": "o1A_ZnuLHvtG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "train_df = pd.DataFrame(train_samples)\n",
        "val_df = pd.DataFrame(val_samples)\n",
        "test_df = pd.DataFrame(test_samples)\n",
        "\n",
        "train_df.head(14)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "zQBRiNV3TKhC",
        "outputId": "0da9db5c-2bb1-4744-eb81-5a8f74cd00f0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         target                                               text  \\\n",
              "0     OBJECTIVE  to investigate the efficacy of @ weeks of dail...   \n",
              "1       METHODS  a total of @ patients with primary knee oa wer...   \n",
              "2       METHODS  outcome measures included pain reduction and i...   \n",
              "3       METHODS  pain was assessed using the visual analog pain...   \n",
              "4       METHODS  secondary outcome measures included the wester...   \n",
              "5       METHODS  serum levels of interleukin @ ( il-@ ) , il-@ ...   \n",
              "6       RESULTS  there was a clinically relevant reduction in t...   \n",
              "7       RESULTS  the mean difference between treatment arms ( @...   \n",
              "8       RESULTS  further , there was a clinically relevant redu...   \n",
              "9       RESULTS  these differences remained significant at @ we...   \n",
              "10      RESULTS  the outcome measures in rheumatology clinical ...   \n",
              "11  CONCLUSIONS  low-dose oral prednisolone had both a short-te...   \n",
              "12   BACKGROUND  emotional eating is associated with overeating...   \n",
              "13   BACKGROUND  yet , empirical evidence for individual ( trai...   \n",
              "\n",
              "    line_number  total_lines  \n",
              "0             0           11  \n",
              "1             1           11  \n",
              "2             2           11  \n",
              "3             3           11  \n",
              "4             4           11  \n",
              "5             5           11  \n",
              "6             6           11  \n",
              "7             7           11  \n",
              "8             8           11  \n",
              "9             9           11  \n",
              "10           10           11  \n",
              "11           11           11  \n",
              "12            0           10  \n",
              "13            1           10  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b2311caf-5ceb-4720-be70-0cc4927fefcd\" 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>target</th>\n",
              "      <th>text</th>\n",
              "      <th>line_number</th>\n",
              "      <th>total_lines</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>OBJECTIVE</td>\n",
              "      <td>to investigate the efficacy of @ weeks of dail...</td>\n",
              "      <td>0</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>a total of @ patients with primary knee oa wer...</td>\n",
              "      <td>1</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>outcome measures included pain reduction and i...</td>\n",
              "      <td>2</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>pain was assessed using the visual analog pain...</td>\n",
              "      <td>3</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>secondary outcome measures included the wester...</td>\n",
              "      <td>4</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>serum levels of interleukin @ ( il-@ ) , il-@ ...</td>\n",
              "      <td>5</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>there was a clinically relevant reduction in t...</td>\n",
              "      <td>6</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the mean difference between treatment arms ( @...</td>\n",
              "      <td>7</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>further , there was a clinically relevant redu...</td>\n",
              "      <td>8</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>these differences remained significant at @ we...</td>\n",
              "      <td>9</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the outcome measures in rheumatology clinical ...</td>\n",
              "      <td>10</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>low-dose oral prednisolone had both a short-te...</td>\n",
              "      <td>11</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>emotional eating is associated with overeating...</td>\n",
              "      <td>0</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>yet , empirical evidence for individual ( trai...</td>\n",
              "      <td>1</td>\n",
              "      <td>10</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-b2311caf-5ceb-4720-be70-0cc4927fefcd')\"\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-b2311caf-5ceb-4720-be70-0cc4927fefcd 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-b2311caf-5ceb-4720-be70-0cc4927fefcd');\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-fcaa21bf-5255-41c9-a430-ae327b867bb7\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-fcaa21bf-5255-41c9-a430-ae327b867bb7')\"\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-fcaa21bf-5255-41c9-a430-ae327b867bb7 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# distribution of labels\n",
        "train_df[\"target\"].value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2VYDkLc9H1ME",
        "outputId": "86412d7a-6c96-4dac-d1d4-740884c8cbe4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "METHODS        59353\n",
              "RESULTS        57953\n",
              "CONCLUSIONS    27168\n",
              "BACKGROUND     21727\n",
              "OBJECTIVE      13839\n",
              "Name: target, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_df[\"total_lines\"].plot.hist()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        },
        "id": "X4BiDOZ5Iwac",
        "outputId": "7fd8b069-7067-44cb-f738-6773a8c7f3fb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: ylabel='Frequency'>"
            ]
          },
          "metadata": {},
          "execution_count": 15
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# get lists of sentences\n",
        "train_sentences = train_df[\"text\"].tolist()\n",
        "val_sentences = val_df[\"text\"].tolist()\n",
        "test_sentences = test_df[\"text\"].tolist()\n",
        "\n",
        "len(train_sentences),len(val_sentences),len(test_sentences)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VAK6JbSTIwW5",
        "outputId": "13e5d399-e396-4d28-a470-bdfe28881f98"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(180040, 30212, 30135)"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Target labels to numbers"
      ],
      "metadata": {
        "id": "58Taqj2vZosG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# one hot encode labels\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "one_hot_encoder = OneHotEncoder(sparse=False) # if true return a sparse matrix else array"
      ],
      "metadata": {
        "id": "-qRBXgEQIwUp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_labels_one_hot = one_hot_encoder.fit_transform(train_df[\"target\"].to_numpy().reshape(-1,1))\n",
        "val_labels_one_hot = one_hot_encoder.transform(val_df[\"target\"].to_numpy().reshape(-1, 1))\n",
        "test_labels_one_hot = one_hot_encoder.transform(test_df[\"target\"].to_numpy().reshape(-1, 1))\n",
        "\n",
        "train_labels_one_hot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2qsDY3wOIwSQ",
        "outputId": "86807ee5-a989-472d-a498-ab7126b43544"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/preprocessing/_encoders.py:868: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., 1., 0.],\n",
              "       [0., 0., 1., 0., 0.],\n",
              "       [0., 0., 1., 0., 0.],\n",
              "       ...,\n",
              "       [0., 0., 0., 0., 1.],\n",
              "       [0., 1., 0., 0., 0.],\n",
              "       [0., 1., 0., 0., 0.]])"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "\n",
        "tf.constant(train_labels_one_hot)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JpWlMdnuIwPh",
        "outputId": "8334eedc-9d18-4e21-d25e-6b985d4a7721"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(180040, 5), dtype=float64, numpy=\n",
              "array([[0., 0., 0., 1., 0.],\n",
              "       [0., 0., 1., 0., 0.],\n",
              "       [0., 0., 1., 0., 0.],\n",
              "       ...,\n",
              "       [0., 0., 0., 0., 1.],\n",
              "       [0., 1., 0., 0., 0.],\n",
              "       [0., 1., 0., 0., 0.]])>"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Label encode labels"
      ],
      "metadata": {
        "id": "POzLwVtZfOqp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import LabelEncoder\n",
        "label_encoder = LabelEncoder()\n",
        "train_labels_encoded = label_encoder.fit_transform(train_df[\"target\"].to_numpy())\n",
        "val_labels_encoded = label_encoder.transform(val_df[\"target\"].to_numpy())\n",
        "test_labels_encoded = label_encoder.transform(test_df[\"target\"].to_numpy())\n",
        "\n",
        "# Check what training labels look like\n",
        "train_labels_encoded"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jtsL3KCAwcmB",
        "outputId": "266bf7d3-8208-42a9-c3af-f2739c02172a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([3, 2, 2, ..., 4, 1, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "num_classes = len(label_encoder.classes_)\n",
        "class_names = label_encoder.classes_\n",
        "\n",
        "num_classes, class_names"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ThEFRlqofgHT",
        "outputId": "ab5fa3f5-d60e-4ae2-f030-75b16b09d7b9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(5,\n",
              " array(['BACKGROUND', 'CONCLUSIONS', 'METHODS', 'OBJECTIVE', 'RESULTS'],\n",
              "       dtype=object))"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Model 0 (Baseline)"
      ],
      "metadata": {
        "id": "vlR6JKvrYceb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.pipeline import Pipeline\n",
        "\n",
        "\n",
        "model_0 = Pipeline([\n",
        "    (\"tfidf\", TfidfVectorizer()),\n",
        "    (\"clf\", MultinomialNB())\n",
        "])\n",
        "\n",
        "model_0.fit(train_sentences, train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 124
        },
        "id": "KxKWYVjxfgDg",
        "outputId": "b4ad9e85-2bf5-45ab-c66c-0371d8b99790"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Pipeline(steps=[('tfidf', TfidfVectorizer()), ('clf', MultinomialNB())])"
            ],
            "text/html": [
              "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;tfidf&#x27;, TfidfVectorizer()), (&#x27;clf&#x27;, MultinomialNB())])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;tfidf&#x27;, TfidfVectorizer()), (&#x27;clf&#x27;, MultinomialNB())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TfidfVectorizer</label><div class=\"sk-toggleable__content\"><pre>TfidfVectorizer()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">MultinomialNB</label><div class=\"sk-toggleable__content\"><pre>MultinomialNB()</pre></div></div></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "baseline_score = model_0.score(val_sentences, val_labels_encoded)\n",
        "print(f\"Our baseline model score is {baseline_score*100:.2f}%\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eLnLnrPafgA_",
        "outputId": "ea13f15b-be76-44be-c7b3-f4ecd603b994"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Our baseline model score is 72.18%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "baseline_preds = model_0.predict(val_sentences)\n",
        "baseline_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PybKtgtAff-a",
        "outputId": "e41c2ca1-03b3-4cd0-be80-9219a6832433"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([4, 1, 3, ..., 4, 4, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Function to calculate results"
      ],
      "metadata": {
        "id": "XbcmnM2sapAW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n",
        "\n",
        "def calculate_results(y_true, y_pred):\n",
        "  model_accuracy = accuracy_score(y_true, y_pred)*100\n",
        "  model_precision, model_recall, model_f1, _ = precision_recall_fscore_support(y_true, y_pred, average=\"weighted\")\n",
        "  model_results = {\"accuracy\": model_accuracy,\n",
        "                  \"precision\": model_precision,\n",
        "                  \"recall\": model_recall,\n",
        "                  \"f1\": model_f1}\n",
        "  return model_results"
      ],
      "metadata": {
        "id": "3h7hZoRyff79"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "baseline_results = calculate_results(val_labels_encoded ,baseline_preds)\n",
        "baseline_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QHKKSKjGff5M",
        "outputId": "7ea42599-5ed6-428b-880b-745e36a5da7d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 72.1832384482987,\n",
              " 'precision': 0.7186466952323352,\n",
              " 'recall': 0.7218323844829869,\n",
              " 'f1': 0.6989250353450294}"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Text vectorization layer"
      ],
      "metadata": {
        "id": "Q6fNWf4_tlb3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "sent_lens = [len(sentence.split()) for sentence in train_sentences]\n",
        "avg_sent_len = np.mean(sent_lens)\n",
        "avg_sent_len"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SpaJ1vkCff2p",
        "outputId": "e3550b03-9679-4b4a-b00c-0b1191e4a401"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "26.338269273494777"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.hist(sent_lens, bins=7);"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "XIUg17Yiff0L",
        "outputId": "33563b26-22ac-4aa4-a9f5-f2338c481bd1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "output_seq_len = int(np.percentile(sent_lens, 95))\n",
        "output_seq_len"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BJzhNp0Affxl",
        "outputId": "296d1ef1-9945-4192-fbeb-a0200ca23345"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "55"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "max(sent_lens)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "N6PWDPM-ffu6",
        "outputId": "05b24de1-c043-4b5f-d276-fd309f485fd8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "296"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating the text vectorizer"
      ],
      "metadata": {
        "id": "r8xh5GAFw53_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# How many words are in our vocabulary? (taken from 3.2 in https://arxiv.org/pdf/1710.06071.pdf)\n",
        "max_tokens = 68000\n",
        "\n",
        "from tensorflow.keras.layers import TextVectorization\n",
        "\n",
        "text_vectorizer = TextVectorization(max_tokens=max_tokens,\n",
        "                                    output_sequence_length=55)"
      ],
      "metadata": {
        "id": "GxfUqUwnw5eq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "text_vectorizer.adapt(train_sentences)"
      ],
      "metadata": {
        "id": "_qh82eADfewq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "target_sentence = random.choice(train_sentences)\n",
        "\n",
        "print(f\"Text:\\n{target_sentence}\")\n",
        "print(f\"Length of text:\\n{len(target_sentence.split())}\")\n",
        "print(f\"\\nVectorized text:\\n{text_vectorizer([target_sentence])}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xNlCGeO0feuO",
        "outputId": "d2490f02-2dd2-48c1-ee76-6c4af8939d04"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Text:\n",
            "implant costs were @ % of initial hospitalization costs .\n",
            "Length of text:\n",
            "10\n",
            "\n",
            "Vectorized text:\n",
            "[[1517  577    9    4  473  843  577    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# how many words in our training vocab\n",
        "\n",
        "rct_20k_text_vocab = text_vectorizer.get_vocabulary()\n",
        "print(f\"Number of words in vocab: {len(rct_20k_text_vocab)}\")\n",
        "print(f\"Most common words in vocab: {rct_20k_text_vocab[:5]}\")\n",
        "print(f\"Most common words in vocab: {rct_20k_text_vocab[-5:]}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t_Tjyrd5feqj",
        "outputId": "4b384c2c-f718-4f2a-f000-86c59700dae8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of words in vocab: 64841\n",
            "Most common words in vocab: ['', '[UNK]', 'the', 'and', 'of']\n",
            "Most common words in vocab: ['aainduced', 'aaigroup', 'aachener', 'aachen', 'aaacp']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "text_vectorizer.get_config()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "834KOkJafeoM",
        "outputId": "22eb6d5c-2b08-4a00-8d1e-5baf5094656a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'name': 'text_vectorization',\n",
              " 'trainable': True,\n",
              " 'dtype': 'string',\n",
              " 'batch_input_shape': (None,),\n",
              " 'max_tokens': 68000,\n",
              " 'standardize': 'lower_and_strip_punctuation',\n",
              " 'split': 'whitespace',\n",
              " 'ngrams': None,\n",
              " 'output_mode': 'int',\n",
              " 'output_sequence_length': 55,\n",
              " 'pad_to_max_tokens': False,\n",
              " 'sparse': False,\n",
              " 'ragged': False,\n",
              " 'vocabulary': None,\n",
              " 'idf_weights': None,\n",
              " 'encoding': 'utf-8',\n",
              " 'vocabulary_size': 64841}"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Creating a token embedding layer"
      ],
      "metadata": {
        "id": "m-wyQyapuJfB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "token_embed = tf.keras.layers.Embedding(input_dim=len(rct_20k_text_vocab),#length of vocab\n",
        "                                        output_dim=128,\n",
        "                                        mask_zero=True,\n",
        "                                        name=\"token_embedding\")\n"
      ],
      "metadata": {
        "id": "MfMfAJMQfelv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(f\"Sentence before vectorization:\\n{target_sentence}\\n\")\n",
        "vectorized_sentence = text_vectorizer([target_sentence])\n",
        "print(f\"Sentence after vectorization (before embedding):\\n{vectorized_sentence}\\n\")\n",
        "embedded_sentence = token_embed(vectorized_sentence)\n",
        "print(f\"Sentence after embedding:\\n{embedded_sentence}\\n\")\n",
        "print(f\"Embedded sentence shape: {embedded_sentence.shape}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "N3Yee1lRfejT",
        "outputId": "2b2667f9-a4c4-4adf-f56c-173a4c46ad22"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Sentence before vectorization:\n",
            "implant costs were @ % of initial hospitalization costs .\n",
            "\n",
            "Sentence after vectorization (before embedding):\n",
            "[[1517  577    9    4  473  843  577    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0    0\n",
            "     0    0    0    0    0    0    0    0    0    0    0    0    0]]\n",
            "\n",
            "Sentence after embedding:\n",
            "[[[-0.03842409 -0.02975982 -0.02932702 ...  0.03394835 -0.03683349\n",
            "   -0.02955512]\n",
            "  [-0.03768674  0.03593886 -0.04171482 ... -0.00396155 -0.02907571\n",
            "   -0.03668419]\n",
            "  [ 0.00717949  0.03172406  0.02992215 ...  0.01460863  0.02373519\n",
            "   -0.0475709 ]\n",
            "  ...\n",
            "  [-0.04551934  0.02602321 -0.01394868 ... -0.03023794  0.00550389\n",
            "    0.02997062]\n",
            "  [-0.04551934  0.02602321 -0.01394868 ... -0.03023794  0.00550389\n",
            "    0.02997062]\n",
            "  [-0.04551934  0.02602321 -0.01394868 ... -0.03023794  0.00550389\n",
            "    0.02997062]]]\n",
            "\n",
            "Embedded sentence shape: (1, 55, 128)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Creating datasets(as fast as possible)"
      ],
      "metadata": {
        "id": "mYRVI0Vmwxg5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_dataset = tf.data.Dataset.from_tensor_slices((train_sentences, train_labels_one_hot))\n",
        "valid_dataset = tf.data.Dataset.from_tensor_slices((val_sentences, val_labels_one_hot))\n",
        "test_dataset = tf.data.Dataset.from_tensor_slices((test_sentences, test_labels_one_hot))\n",
        "\n",
        "train_dataset\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sU4Nhl2Qfeg5",
        "outputId": "43768867-2fdb-4deb-f114-86cf40371a06"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<_TensorSliceDataset element_spec=(TensorSpec(shape=(), dtype=tf.string, name=None), TensorSpec(shape=(5,), dtype=tf.float64, name=None))>"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# TensorSliceDataset converting to prefected datasets\n",
        "train_dataset = train_dataset.batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "valid_dataset = valid_dataset.batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "test_dataset = test_dataset.batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "\n",
        "train_dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ak-ZAUEo0gOO",
        "outputId": "9980401f-02db-4afd-9dfd-2bf2c8dad161"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<_PrefetchDataset element_spec=(TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model-1 (Conv1D)"
      ],
      "metadata": {
        "id": "9dHKmZ9K8g2G"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "inputs = tf.keras.layers.Input(shape=(1,), dtype=tf.string)\n",
        "text_vectors = text_vectorizer(inputs)\n",
        "token_embeddings = token_embed(text_vectors)\n",
        "x = tf.keras.layers.Conv1D(64, kernel_size=5, padding=\"same\", activation=\"relu\")(token_embeddings)\n",
        "x = tf.keras.layers.GlobalAveragePooling1D()(x)\n",
        "\n",
        "outputs = tf.keras.layers.Dense(num_classes, activation=\"softmax\")(x)\n",
        "\n",
        "model_1 = tf.keras.Model(inputs,outputs, name=\"model_1_conv1D\")\n",
        "\n",
        "model_1.compile(loss=\"categorical_crossentropy\", # if your labels are integer form (not one hot) use sparse_categorical_crossentropy\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "IzQfZgui8EEH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ozJdv_KE_cW4",
        "outputId": "28ab60cc-2405-46b8-f3ba-ed9efb6230fd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_1_conv1D\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 1)]               0         \n",
            "                                                                 \n",
            " text_vectorization (TextVe  (None, 55)                0         \n",
            " ctorization)                                                    \n",
            "                                                                 \n",
            " token_embedding (Embedding  (None, 55, 128)           8299648   \n",
            " )                                                               \n",
            "                                                                 \n",
            " conv1d (Conv1D)             (None, 55, 64)            41024     \n",
            "                                                                 \n",
            " global_average_pooling1d (  (None, 64)                0         \n",
            " GlobalAveragePooling1D)                                         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 8340997 (31.82 MB)\n",
            "Trainable params: 8340997 (31.82 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1_history = model_1.fit(train_dataset,\n",
        "                              steps_per_epoch=int(0.1 * len(train_dataset)), # only fit on 10% of batches for faster training time\n",
        "                              epochs=3,\n",
        "                              validation_data=valid_dataset,\n",
        "                              validation_steps=int(0.1 * len(valid_dataset)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fFACGs4E_eV1",
        "outputId": "c4fb1342-18a0-499e-abcb-946a258a43d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "562/562 [==============================] - 21s 33ms/step - loss: 0.9199 - accuracy: 0.6380 - val_loss: 0.6909 - val_accuracy: 0.7357\n",
            "Epoch 2/3\n",
            "562/562 [==============================] - 5s 9ms/step - loss: 0.6607 - accuracy: 0.7546 - val_loss: 0.6345 - val_accuracy: 0.7683\n",
            "Epoch 3/3\n",
            "562/562 [==============================] - 5s 9ms/step - loss: 0.6159 - accuracy: 0.7761 - val_loss: 0.5963 - val_accuracy: 0.7859\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.evaluate(valid_dataset)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5q2iAjdD_yVt",
        "outputId": "efe24b9a-3683-466d-e2aa-ebe42f82d143"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 4s 4ms/step - loss: 0.5969 - accuracy: 0.7881\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.5969167947769165, 0.7880974411964417]"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1_pred_probs = model_1.predict(valid_dataset)\n",
        "model_1_pred_probs"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KyrFuAF2AAN_",
        "outputId": "b3ae82e6-1929-438a-9241-787a4b995d4e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 2s 2ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[4.7678253e-01, 1.5140826e-01, 5.0389037e-02, 3.0217141e-01,\n",
              "        1.9248687e-02],\n",
              "       [3.9293879e-01, 3.0382746e-01, 1.2106692e-02, 2.8209507e-01,\n",
              "        9.0319654e-03],\n",
              "       [1.5591168e-01, 7.2839269e-03, 2.1826038e-03, 8.3458930e-01,\n",
              "        3.2419426e-05],\n",
              "       ...,\n",
              "       [1.0008975e-05, 8.1598107e-04, 1.0182167e-03, 4.8235825e-06,\n",
              "        9.9815089e-01],\n",
              "       [4.4962808e-02, 4.7277087e-01, 8.0044605e-02, 5.7850856e-02,\n",
              "        3.4437093e-01],\n",
              "       [1.6994941e-01, 6.9417936e-01, 4.5385621e-02, 4.7686819e-02,\n",
              "        4.2798772e-02]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1_preds = tf.argmax(model_1_pred_probs, axis=1)\n",
        "model_1_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1yQ1P0NjAGMZ",
        "outputId": "21ed6509-14e3-43a7-fb4d-cc262bbee30d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(30212,), dtype=int64, numpy=array([0, 0, 3, ..., 4, 1, 1])>"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1_results = calculate_results(val_labels_encoded,\n",
        "                                    model_1_preds)\n",
        "model_1_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nmkLkbd9AKBE",
        "outputId": "2c33f9d0-c6db-42f5-b14c-b78e6781f4a3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 78.80974447239507,\n",
              " 'precision': 0.7850363583625222,\n",
              " 'recall': 0.7880974447239507,\n",
              " 'f1': 0.7856801315289836}"
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Model 2: Feature extraction with pretrained token embeddings\n"
      ],
      "metadata": {
        "id": "vBm31IdgOEIP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow_hub as hub\n",
        "tf_hub_embedding_layer = hub.KerasLayer(\"https://tfhub.dev/google/universal-sentence-encoder/4\",\n",
        "                                        trainable=False,\n",
        "                                        name=\"universal_sentence_encoder\")"
      ],
      "metadata": {
        "id": "XyvjKrFbAYkm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "random_training_sentence = random.choice(train_sentences)\n",
        "print(f\"Random training sentence:\\n{random_training_sentence}\\n\")\n",
        "use_embedded_sentence = tf_hub_embedding_layer([random_training_sentence])\n",
        "print(f\"Sentence after embedding:\\n{use_embedded_sentence[0][:30]} (truncated output)...\\n\")\n",
        "print(f\"Length of sentence embedding:\\n{len(use_embedded_sentence[0])}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lLFuITulOeAF",
        "outputId": "7ca213d0-5176-4539-87af-cd591af266df"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Random training sentence:\n",
            "in the highest baseline bmi group , the mean increase in triglycerides was @ and @ % in the atv/r and lpv/r arms , respectively .\n",
            "\n",
            "Sentence after embedding:\n",
            "[-0.00971456  0.01037106 -0.0104351  -0.0273001  -0.02249291 -0.01510036\n",
            " -0.03056139 -0.05113742 -0.05349253  0.04678805  0.0636134   0.01874467\n",
            " -0.06427366 -0.02573305 -0.06443615 -0.00749563 -0.08574142 -0.00611815\n",
            " -0.01280435  0.04823345  0.03451956  0.00217599 -0.01295516 -0.04995313\n",
            "  0.02335102  0.04106525  0.01063864  0.02472979  0.02218883 -0.06373401] (truncated output)...\n",
            "\n",
            "Length of sentence embedding:\n",
            "512\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "inputs = tf.keras.layers.Input(shape=[], dtype=tf.string)\n",
        "pretrained_embedding = tf_hub_embedding_layer(inputs) # tokenize text and create embedding\n",
        "x = tf.keras.layers.Dense(128, activation=\"relu\")(pretrained_embedding) # add a fully connected layer on top of the embedding\n",
        "outputs = tf.keras.layers.Dense(5, activation=\"softmax\")(x) # create the output layer\n",
        "model_2 = tf.keras.Model(inputs=inputs,\n",
        "                        outputs=outputs)\n",
        "\n",
        "# Compile the model\n",
        "model_2.compile(loss=\"categorical_crossentropy\",\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "4QFaML1wOiD8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.summary()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "G6wU_VhAO4-V",
        "outputId": "f8e7b072-aaab-49c2-a2bf-3c42b91f1cc7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None,)]                 0         \n",
            "                                                                 \n",
            " universal_sentence_encoder  (None, 512)               256797824 \n",
            "  (KerasLayer)                                                   \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 128)               65664     \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 5)                 645       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 256864133 (979.86 MB)\n",
            "Trainable params: 66309 (259.02 KB)\n",
            "Non-trainable params: 256797824 (979.61 MB)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.fit(train_dataset,\n",
        "            steps_per_epoch=int(0.1 * len(train_dataset)),\n",
        "            epochs=3,\n",
        "            validation_data=valid_dataset,\n",
        "            validation_steps=int(0.1 * len(valid_dataset)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EiHN4RNgO-CB",
        "outputId": "f61f534c-e223-401c-cbf7-01ccac959fa0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "562/562 [==============================] - 14s 20ms/step - loss: 0.9177 - accuracy: 0.6492 - val_loss: 0.7973 - val_accuracy: 0.6872\n",
            "Epoch 2/3\n",
            "562/562 [==============================] - 7s 13ms/step - loss: 0.7702 - accuracy: 0.7013 - val_loss: 0.7567 - val_accuracy: 0.7064\n",
            "Epoch 3/3\n",
            "562/562 [==============================] - 8s 14ms/step - loss: 0.7554 - accuracy: 0.7092 - val_loss: 0.7424 - val_accuracy: 0.7124\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7942665e07f0>"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.evaluate(valid_dataset)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TPEzBVYcO_ri",
        "outputId": "8d8c3ca9-37ed-4a87-a063-67f4273b916f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 11s 12ms/step - loss: 0.7443 - accuracy: 0.7113\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.7442595362663269, 0.7113067507743835]"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2_pred_probs = model_2.predict(valid_dataset)\n",
        "model_2_pred_probs"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LHsdGFRxPCaT",
        "outputId": "742b0f22-bb91-4dce-cf7b-11c77e7879c6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 9s 9ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.43948236, 0.3484731 , 0.00257211, 0.20166351, 0.00780899],\n",
              "       [0.30325502, 0.54680204, 0.00290063, 0.14422889, 0.00281338],\n",
              "       [0.20376584, 0.13518001, 0.02208353, 0.6028599 , 0.03611076],\n",
              "       ...,\n",
              "       [0.00145318, 0.00738455, 0.06176007, 0.00100198, 0.9284003 ],\n",
              "       [0.0032596 , 0.04264791, 0.19876996, 0.00161637, 0.75370616],\n",
              "       [0.15202355, 0.22785135, 0.54657817, 0.0079795 , 0.06556752]],\n",
              "      dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2_preds = tf.argmax(model_2_pred_probs, axis=1)\n",
        "model_2_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FJhuQ10JPJCe",
        "outputId": "261670a1-3efb-45a0-defb-c46d8bddb5be"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(30212,), dtype=int64, numpy=array([0, 1, 3, ..., 4, 4, 2])>"
            ]
          },
          "metadata": {},
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2_results = calculate_results(y_true=val_labels_encoded,\n",
        "                                    y_pred=model_2_preds)\n",
        "model_2_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lNpZ_HzKPKf_",
        "outputId": "99ee62a4-182f-4a83-c9fe-a33f4de3c48c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 71.1306765523633,\n",
              " 'precision': 0.711305286242712,\n",
              " 'recall': 0.711306765523633,\n",
              " 'f1': 0.7081875393244097}"
            ]
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Model 3: Conv1D with character embeddings"
      ],
      "metadata": {
        "id": "iv5hJzeaPRG2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating a character-level tokenizer"
      ],
      "metadata": {
        "id": "SCXVrT5v5fhq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_sentences[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8S1LCfrLPMAy",
        "outputId": "f55d98f3-9582-4342-f589-48c8f26b0763"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['to investigate the efficacy of @ weeks of daily low-dose oral prednisolone in improving pain , mobility , and systemic low-grade inflammation in the short term and whether the effect would be sustained at @ weeks in older adults with moderate to severe knee osteoarthritis ( oa ) .',\n",
              " 'a total of @ patients with primary knee oa were randomized @:@ ; @ received @ mg/day of prednisolone and @ received placebo for @ weeks .',\n",
              " 'outcome measures included pain reduction and improvement in function scores and systemic inflammation markers .',\n",
              " 'pain was assessed using the visual analog pain scale ( @-@ mm ) .',\n",
              " 'secondary outcome measures included the western ontario and mcmaster universities osteoarthritis index scores , patient global assessment ( pga ) of the severity of knee oa , and @-min walk distance ( @mwd ) .']"
            ]
          },
          "metadata": {},
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def split_chars(text):\n",
        "  return \" \".join(list(text))\n",
        "\n",
        "split_chars(random_training_sentence)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "id": "ErjgH6LkPU6y",
        "outputId": "19572d18-4ede-4343-b2a4-dceafa495689"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'i n   t h e   h i g h e s t   b a s e l i n e   b m i   g r o u p   ,   t h e   m e a n   i n c r e a s e   i n   t r i g l y c e r i d e s   w a s   @   a n d   @   %   i n   t h e   a t v / r   a n d   l p v / r   a r m s   ,   r e s p e c t i v e l y   .'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_chars = [split_chars(sentence) for sentence in train_sentences]\n",
        "val_chars = [split_chars(sentence) for sentence in val_sentences]\n",
        "test_chars = [split_chars(sentence) for sentence in test_sentences]\n",
        "train_chars[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fYhsQBB7PU3J",
        "outputId": "ca66a498-d749-473b-f93f-b0a2e5bd3084"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['t o   i n v e s t i g a t e   t h e   e f f i c a c y   o f   @   w e e k s   o f   d a i l y   l o w - d o s e   o r a l   p r e d n i s o l o n e   i n   i m p r o v i n g   p a i n   ,   m o b i l i t y   ,   a n d   s y s t e m i c   l o w - g r a d e   i n f l a m m a t i o n   i n   t h e   s h o r t   t e r m   a n d   w h e t h e r   t h e   e f f e c t   w o u l d   b e   s u s t a i n e d   a t   @   w e e k s   i n   o l d e r   a d u l t s   w i t h   m o d e r a t e   t o   s e v e r e   k n e e   o s t e o a r t h r i t i s   (   o a   )   .',\n",
              " 'a   t o t a l   o f   @   p a t i e n t s   w i t h   p r i m a r y   k n e e   o a   w e r e   r a n d o m i z e d   @ : @   ;   @   r e c e i v e d   @   m g / d a y   o f   p r e d n i s o l o n e   a n d   @   r e c e i v e d   p l a c e b o   f o r   @   w e e k s   .',\n",
              " 'o u t c o m e   m e a s u r e s   i n c l u d e d   p a i n   r e d u c t i o n   a n d   i m p r o v e m e n t   i n   f u n c t i o n   s c o r e s   a n d   s y s t e m i c   i n f l a m m a t i o n   m a r k e r s   .',\n",
              " 'p a i n   w a s   a s s e s s e d   u s i n g   t h e   v i s u a l   a n a l o g   p a i n   s c a l e   (   @ - @   m m   )   .',\n",
              " 's e c o n d a r y   o u t c o m e   m e a s u r e s   i n c l u d e d   t h e   w e s t e r n   o n t a r i o   a n d   m c m a s t e r   u n i v e r s i t i e s   o s t e o a r t h r i t i s   i n d e x   s c o r e s   ,   p a t i e n t   g l o b a l   a s s e s s m e n t   (   p g a   )   o f   t h e   s e v e r i t y   o f   k n e e   o a   ,   a n d   @ - m i n   w a l k   d i s t a n c e   (   @ m w d   )   .']"
            ]
          },
          "metadata": {},
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# avg character length?\n",
        "char_lens = [len(sentence) for sentence in train_sentences]\n",
        "mean_char_len = np.mean(char_lens)\n",
        "mean_char_len"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NJ_Okny1PUzY",
        "outputId": "9c4de4ee-7bb3-4c58-a3e8-0273fb337b6b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "149.3662574983337"
            ]
          },
          "metadata": {},
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# check the distribution of our sequences at a character-level\n",
        "import matplotlib.pyplot as plt\n",
        "plt.hist(char_lens, bins=7)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 517
        },
        "id": "aCStUssVPUvn",
        "outputId": "1ee1cfb4-4c29-4d7f-dde1-5f4fae517319"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array([1.41175e+05, 3.71110e+04, 1.60000e+03, 1.27000e+02, 2.10000e+01,\n",
              "        5.00000e+00, 1.00000e+00]),\n",
              " array([1.00000000e+00, 1.98857143e+02, 3.96714286e+02, 5.94571429e+02,\n",
              "        7.92428571e+02, 9.90285714e+02, 1.18814286e+03, 1.38600000e+03]),\n",
              " <BarContainer object of 7 artists>)"
            ]
          },
          "metadata": {},
          "execution_count": 60
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkIAAAGdCAYAAAD+JxxnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAz6UlEQVR4nO3df3BU9b3/8VdCyA/Q3fCjybI1QG7L5UdJQYmEINI67BBLtDeVtgRTpJrC1SZKDPJLMGKrDcZrBfxBSntbmCkUZEZSDRhMgxKVGCAQIUginSKgdBP7DdmVKBDI+f7h5FwWEMRuiMnn+Zg5M+75vM/nfD6fMdnXnJxzCLEsyxIAAICBQjt6AAAAAB2FIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMFZYRw/g66y1tVXHjh3Ttddeq5CQkI4eDgAA+BIsy9Inn3wit9ut0NBLX/MhCF3CsWPHFBcX19HDAAAAX8HRo0d13XXXXbKGIHQJ1157raTPF9LhcHTwaAAAwJfh9/sVFxdnf49fCkHoEtr+HOZwOAhCAAB0Ml/mthZulgYAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgrCsOQuXl5br99tvldrsVEhKioqKiL6y99957FRISoqVLlwbsb2xsVEZGhhwOh6Kjo5WZmakTJ04E1Ozdu1c333yzIiMjFRcXp4KCggv637Bhg4YMGaLIyEglJCRo8+bNAe2WZSkvL0/9+vVTVFSUPB6PDh48eKVTBgAAXVTYlR7Q3NysESNG6J577tEdd9zxhXUbN27UO++8I7fbfUFbRkaG/vnPf6q0tFQtLS26++67NXPmTK1du1aS5Pf7NXHiRHk8HhUWFmrfvn265557FB0drZkzZ0qStm/frqlTpyo/P1+33Xab1q5dq7S0NO3evVvDhw+XJBUUFGj58uVavXq14uPj9cgjjyglJUXvvfeeIiMjr3TqQTdw/qaOHkKH+mBJakcPAQBguBDLsqyvfHBIiDZu3Ki0tLSA/R999JGSkpK0ZcsWpaamKicnRzk5OZKkAwcOaNiwYdq5c6cSExMlSSUlJZo0aZI+/PBDud1urVixQgsXLpTX61V4eLgkaf78+SoqKlJtba0kacqUKWpublZxcbF93jFjxmjkyJEqLCyUZVlyu92aPXu2HnroIUmSz+dTbGysVq1apfT09MvOz+/3y+l0yufzyeFwfNVl+kIEIYIQACD4ruT7O+j3CLW2tmratGmaM2eOvvOd71zQXlFRoejoaDsESZLH41FoaKgqKyvtmvHjx9shSJJSUlJUV1en48eP2zUejyeg75SUFFVUVEiSDh06JK/XG1DjdDqVlJRk15zv1KlT8vv9ARsAAOi6gh6EnnzySYWFhemBBx64aLvX61VMTEzAvrCwMPXu3Vter9euiY2NDahp+3y5mnPbzz3uYjXny8/Pl9PptLe4uLjLzhcAAHReQQ1CVVVVWrZsmVatWqWQkJBgdn1VLFiwQD6fz96OHj3a0UMCAADtKKhB6M0331RDQ4P69++vsLAwhYWF6fDhw5o9e7YGDhwoSXK5XGpoaAg47syZM2psbJTL5bJr6uvrA2raPl+u5tz2c4+7WM35IiIi5HA4AjYAANB1BTUITZs2TXv37lV1dbW9ud1uzZkzR1u2bJEkJScnq6mpSVVVVfZxW7duVWtrq5KSkuya8vJytbS02DWlpaUaPHiwevXqZdeUlZUFnL+0tFTJycmSpPj4eLlcroAav9+vyspKuwYAAJjtih+fP3HihP7+97/bnw8dOqTq6mr17t1b/fv3V58+fQLqu3fvLpfLpcGDB0uShg4dqltvvVUzZsxQYWGhWlpalJ2drfT0dPtR+zvvvFOPPfaYMjMzNW/ePNXU1GjZsmV65pln7H5nzZql733ve3r66aeVmpqqdevWadeuXVq5cqWkz59oy8nJ0eOPP65BgwbZj8+73e4LnnIDAABmuuIgtGvXLt1yyy3259zcXEnS9OnTtWrVqi/Vx5o1a5Sdna0JEyYoNDRUkydP1vLly+12p9Op1157TVlZWRo1apT69u2rvLw8+x1CkjR27FitXbtWixYt0sMPP6xBgwapqKjIfoeQJM2dO1fNzc2aOXOmmpqaNG7cOJWUlHwt3iEEAAA63r/1HqGujvcItS/eIwQAaA8d+h4hAACAzoIgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGCsKw5C5eXluv322+V2uxUSEqKioiK7raWlRfPmzVNCQoJ69uwpt9utu+66S8eOHQvoo7GxURkZGXI4HIqOjlZmZqZOnDgRULN3717dfPPNioyMVFxcnAoKCi4Yy4YNGzRkyBBFRkYqISFBmzdvDmi3LEt5eXnq16+foqKi5PF4dPDgwSudMgAA6KKuOAg1NzdrxIgRev755y9o+/TTT7V792498sgj2r17t1566SXV1dXphz/8YUBdRkaG9u/fr9LSUhUXF6u8vFwzZ8602/1+vyZOnKgBAwaoqqpKTz31lBYvXqyVK1faNdu3b9fUqVOVmZmpPXv2KC0tTWlpaaqpqbFrCgoKtHz5chUWFqqyslI9e/ZUSkqKTp48eaXTBgAAXVCIZVnWVz44JEQbN25UWlraF9bs3LlTo0eP1uHDh9W/f38dOHBAw4YN086dO5WYmChJKikp0aRJk/Thhx/K7XZrxYoVWrhwobxer8LDwyVJ8+fPV1FRkWprayVJU6ZMUXNzs4qLi+1zjRkzRiNHjlRhYaEsy5Lb7dbs2bP10EMPSZJ8Pp9iY2O1atUqpaenX3Z+fr9fTqdTPp9PDofjqy7TFxo4f1PQ++xMPliS2tFDAAB0QVfy/d3u9wj5fD6FhIQoOjpaklRRUaHo6Gg7BEmSx+NRaGioKisr7Zrx48fbIUiSUlJSVFdXp+PHj9s1Ho8n4FwpKSmqqKiQJB06dEherzegxul0Kikpya4536lTp+T3+wM2AADQdbVrEDp58qTmzZunqVOn2onM6/UqJiYmoC4sLEy9e/eW1+u1a2JjYwNq2j5frubc9nOPu1jN+fLz8+V0Ou0tLi7uiucMAAA6j3YLQi0tLfrpT38qy7K0YsWK9jpNUC1YsEA+n8/ejh492tFDAgAA7SisPTptC0GHDx/W1q1bA/4+53K51NDQEFB/5swZNTY2yuVy2TX19fUBNW2fL1dzbnvbvn79+gXUjBw58qLjjoiIUERExJVOFwAAdFJBvyLUFoIOHjyov/3tb+rTp09Ae3JyspqamlRVVWXv27p1q1pbW5WUlGTXlJeXq6Wlxa4pLS3V4MGD1atXL7umrKwsoO/S0lIlJydLkuLj4+VyuQJq/H6/Kisr7RoAAGC2Kw5CJ06cUHV1taqrqyV9flNydXW1jhw5opaWFv34xz/Wrl27tGbNGp09e1Zer1der1enT5+WJA0dOlS33nqrZsyYoR07dujtt99Wdna20tPT5Xa7JUl33nmnwsPDlZmZqf3792v9+vVatmyZcnNz7XHMmjVLJSUlevrpp1VbW6vFixdr165dys7OlvT5E205OTl6/PHH9fLLL2vfvn2666675Ha7L/mUGwAAMMcVPz7/xhtv6JZbbrlg//Tp07V48WLFx8df9LjXX39d3//+9yV9/kLF7OxsvfLKKwoNDdXkyZO1fPlyXXPNNXb93r17lZWVpZ07d6pv3766//77NW/evIA+N2zYoEWLFumDDz7QoEGDVFBQoEmTJtntlmXp0Ucf1cqVK9XU1KRx48bphRde0H/+539+qbny+Hz74vF5AEB7uJLv73/rPUJdHUGofRGEAADt4Wv1HiEAAICvK4IQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLGuOAiVl5fr9ttvl9vtVkhIiIqKigLaLctSXl6e+vXrp6ioKHk8Hh08eDCgprGxURkZGXI4HIqOjlZmZqZOnDgRULN3717dfPPNioyMVFxcnAoKCi4Yy4YNGzRkyBBFRkYqISFBmzdvvuKxAAAAc11xEGpubtaIESP0/PPPX7S9oKBAy5cvV2FhoSorK9WzZ0+lpKTo5MmTdk1GRob279+v0tJSFRcXq7y8XDNnzrTb/X6/Jk6cqAEDBqiqqkpPPfWUFi9erJUrV9o127dv19SpU5WZmak9e/YoLS1NaWlpqqmpuaKxAAAAc4VYlmV95YNDQrRx40alpaVJ+vwKjNvt1uzZs/XQQw9Jknw+n2JjY7Vq1Sqlp6frwIEDGjZsmHbu3KnExERJUklJiSZNmqQPP/xQbrdbK1as0MKFC+X1ehUeHi5Jmj9/voqKilRbWytJmjJlipqbm1VcXGyPZ8yYMRo5cqQKCwu/1Fgux+/3y+l0yufzyeFwfNVl+kID528Kep+dyQdLUjt6CACALuhKvr+Deo/QoUOH5PV65fF47H1Op1NJSUmqqKiQJFVUVCg6OtoOQZLk8XgUGhqqyspKu2b8+PF2CJKklJQU1dXV6fjx43bNuedpq2k7z5cZy/lOnTolv98fsAEAgK4rqEHI6/VKkmJjYwP2x8bG2m1er1cxMTEB7WFhYerdu3dAzcX6OPccX1RzbvvlxnK+/Px8OZ1Oe4uLi/sSswYAAJ0VT42dY8GCBfL5fPZ29OjRjh4SAABoR0ENQi6XS5JUX18fsL++vt5uc7lcamhoCGg/c+aMGhsbA2ou1se55/iimnPbLzeW80VERMjhcARsAACg6wpqEIqPj5fL5VJZWZm9z+/3q7KyUsnJyZKk5ORkNTU1qaqqyq7ZunWrWltblZSUZNeUl5erpaXFriktLdXgwYPVq1cvu+bc87TVtJ3ny4wFAACY7YqD0IkTJ1RdXa3q6mpJn9+UXF1drSNHjigkJEQ5OTl6/PHH9fLLL2vfvn2666675Ha77SfLhg4dqltvvVUzZszQjh079Pbbbys7O1vp6elyu92SpDvvvFPh4eHKzMzU/v37tX79ei1btky5ubn2OGbNmqWSkhI9/fTTqq2t1eLFi7Vr1y5lZ2dL0pcaCwAAMFvYlR6wa9cu3XLLLfbntnAyffp0rVq1SnPnzlVzc7NmzpyppqYmjRs3TiUlJYqMjLSPWbNmjbKzszVhwgSFhoZq8uTJWr58ud3udDr12muvKSsrS6NGjVLfvn2Vl5cX8K6hsWPHau3atVq0aJEefvhhDRo0SEVFRRo+fLhd82XGAgAAzPVvvUeoq+M9Qu2L9wgBANpDh71HCAAAoDMhCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYK+hB6OzZs3rkkUcUHx+vqKgofetb39Kvf/1rWZZl11iWpby8PPXr109RUVHyeDw6ePBgQD+NjY3KyMiQw+FQdHS0MjMzdeLEiYCavXv36uabb1ZkZKTi4uJUUFBwwXg2bNigIUOGKDIyUgkJCdq8eXOwpwwAADqpoAehJ598UitWrNBzzz2nAwcO6Mknn1RBQYGeffZZu6agoEDLly9XYWGhKisr1bNnT6WkpOjkyZN2TUZGhvbv36/S0lIVFxervLxcM2fOtNv9fr8mTpyoAQMGqKqqSk899ZQWL16slStX2jXbt2/X1KlTlZmZqT179igtLU1paWmqqakJ9rQBAEAnFGKde6kmCG677TbFxsbqf//3f+19kydPVlRUlP785z/Lsiy53W7Nnj1bDz30kCTJ5/MpNjZWq1atUnp6ug4cOKBhw4Zp586dSkxMlCSVlJRo0qRJ+vDDD+V2u7VixQotXLhQXq9X4eHhkqT58+erqKhItbW1kqQpU6aoublZxcXF9ljGjBmjkSNHqrCw8LJz8fv9cjqd8vl8cjgcQVujNgPnbwp6n53JB0tSO3oIAIAu6Eq+v4N+RWjs2LEqKyvT+++/L0l699139dZbb+kHP/iBJOnQoUPyer3yeDz2MU6nU0lJSaqoqJAkVVRUKDo62g5BkuTxeBQaGqrKykq7Zvz48XYIkqSUlBTV1dXp+PHjds2552mraTvP+U6dOiW/3x+wAQCAriss2B3Onz9ffr9fQ4YMUbdu3XT27Fk98cQTysjIkCR5vV5JUmxsbMBxsbGxdpvX61VMTEzgQMPC1Lt374Ca+Pj4C/poa+vVq5e8Xu8lz3O+/Px8PfbYY19l2gAAoBMK+hWhF198UWvWrNHatWu1e/durV69Wv/zP/+j1atXB/tUQbdgwQL5fD57O3r0aEcPCQAAtKOgXxGaM2eO5s+fr/T0dElSQkKCDh8+rPz8fE2fPl0ul0uSVF9fr379+tnH1dfXa+TIkZIkl8ulhoaGgH7PnDmjxsZG+3iXy6X6+vqAmrbPl6tpaz9fRESEIiIivsq0AQBAJxT0K0KffvqpQkMDu+3WrZtaW1slSfHx8XK5XCorK7Pb/X6/KisrlZycLElKTk5WU1OTqqqq7JqtW7eqtbVVSUlJdk15eblaWlrsmtLSUg0ePFi9evWya849T1tN23kAAIDZgh6Ebr/9dj3xxBPatGmTPvjgA23cuFG//e1v9aMf/UiSFBISopycHD3++ON6+eWXtW/fPt11111yu91KS0uTJA0dOlS33nqrZsyYoR07dujtt99Wdna20tPT5Xa7JUl33nmnwsPDlZmZqf3792v9+vVatmyZcnNz7bHMmjVLJSUlevrpp1VbW6vFixdr165dys7ODva0AQBAJxT0P409++yzeuSRR/TLX/5SDQ0Ncrvd+u///m/l5eXZNXPnzlVzc7NmzpyppqYmjRs3TiUlJYqMjLRr1qxZo+zsbE2YMEGhoaGaPHmyli9fbrc7nU699tprysrK0qhRo9S3b1/l5eUFvGto7NixWrt2rRYtWqSHH35YgwYNUlFRkYYPHx7saQMAgE4o6O8R6kp4j1D74j1CAID20KHvEQIAAOgsCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjNUuQeijjz7Sz372M/Xp00dRUVFKSEjQrl277HbLspSXl6d+/fopKipKHo9HBw8eDOijsbFRGRkZcjgcio6OVmZmpk6cOBFQs3fvXt18882KjIxUXFycCgoKLhjLhg0bNGTIEEVGRiohIUGbN29ujykDAIBOKOhB6Pjx47rpppvUvXt3vfrqq3rvvff09NNPq1evXnZNQUGBli9frsLCQlVWVqpnz55KSUnRyZMn7ZqMjAzt379fpaWlKi4uVnl5uWbOnGm3+/1+TZw4UQMGDFBVVZWeeuopLV68WCtXrrRrtm/frqlTpyozM1N79uxRWlqa0tLSVFNTE+xpAwCATijEsiwrmB3Onz9fb7/9tt58882LtluWJbfbrdmzZ+uhhx6SJPl8PsXGxmrVqlVKT0/XgQMHNGzYMO3cuVOJiYmSpJKSEk2aNEkffvih3G63VqxYoYULF8rr9So8PNw+d1FRkWprayVJU6ZMUXNzs4qLi+3zjxkzRiNHjlRhYeFl5+L3++V0OuXz+eRwOP6tdbmYgfM3Bb3PzuSDJakdPQQAQBd0Jd/fQb8i9PLLLysxMVE/+clPFBMTo+uvv16///3v7fZDhw7J6/XK4/HY+5xOp5KSklRRUSFJqqioUHR0tB2CJMnj8Sg0NFSVlZV2zfjx4+0QJEkpKSmqq6vT8ePH7Zpzz9NW03YeAABgtqAHoX/84x9asWKFBg0apC1btui+++7TAw88oNWrV0uSvF6vJCk2NjbguNjYWLvN6/UqJiYmoD0sLEy9e/cOqLlYH+ee44tq2trPd+rUKfn9/oANAAB0XWHB7rC1tVWJiYn6zW9+I0m6/vrrVVNTo8LCQk2fPj3Ypwuq/Px8PfbYYx09DAAAcJUE/YpQv379NGzYsIB9Q4cO1ZEjRyRJLpdLklRfXx9QU19fb7e5XC41NDQEtJ85c0aNjY0BNRfr49xzfFFNW/v5FixYIJ/PZ29Hjx79cpMGAACdUtCD0E033aS6urqAfe+//74GDBggSYqPj5fL5VJZWZnd7vf7VVlZqeTkZElScnKympqaVFVVZdds3bpVra2tSkpKsmvKy8vV0tJi15SWlmrw4MH2E2rJyckB52mraTvP+SIiIuRwOAI2AADQdQU9CD344IN655139Jvf/EZ///vftXbtWq1cuVJZWVmSpJCQEOXk5Ojxxx/Xyy+/rH379umuu+6S2+1WWlqapM+vIN16662aMWOGduzYobffflvZ2dlKT0+X2+2WJN15550KDw9XZmam9u/fr/Xr12vZsmXKzc21xzJr1iyVlJTo6aefVm1trRYvXqxdu3YpOzs72NMGAACdUNDvEbrxxhu1ceNGLViwQL/61a8UHx+vpUuXKiMjw66ZO3eumpubNXPmTDU1NWncuHEqKSlRZGSkXbNmzRplZ2drwoQJCg0N1eTJk7V8+XK73el06rXXXlNWVpZGjRqlvn37Ki8vL+BdQ2PHjtXatWu1aNEiPfzwwxo0aJCKioo0fPjwYE8bAAB0QkF/j1BXwnuE2hfvEQIAtIcOfY8QAABAZ0EQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADBWuwehJUuWKCQkRDk5Ofa+kydPKisrS3369NE111yjyZMnq76+PuC4I0eOKDU1VT169FBMTIzmzJmjM2fOBNS88cYbuuGGGxQREaFvf/vbWrVq1QXnf/755zVw4EBFRkYqKSlJO3bsaI9pAgCATqhdg9DOnTv1u9/9Tt/97ncD9j/44IN65ZVXtGHDBm3btk3Hjh3THXfcYbefPXtWqampOn36tLZv367Vq1dr1apVysvLs2sOHTqk1NRU3XLLLaqurlZOTo5+8YtfaMuWLXbN+vXrlZubq0cffVS7d+/WiBEjlJKSooaGhvacNgAA6CRCLMuy2qPjEydO6IYbbtALL7ygxx9/XCNHjtTSpUvl8/n0jW98Q2vXrtWPf/xjSVJtba2GDh2qiooKjRkzRq+++qpuu+02HTt2TLGxsZKkwsJCzZs3Tx9//LHCw8M1b948bdq0STU1NfY509PT1dTUpJKSEklSUlKSbrzxRj333HOSpNbWVsXFxen+++/X/PnzLzsHv98vp9Mpn88nh8MR7CXSwPmbgt5nZ/LBktSOHgIAoAu6ku/vdrsilJWVpdTUVHk8noD9VVVVamlpCdg/ZMgQ9e/fXxUVFZKkiooKJSQk2CFIklJSUuT3+7V//3675vy+U1JS7D5Onz6tqqqqgJrQ0FB5PB67BgAAmC2sPTpdt26ddu/erZ07d17Q5vV6FR4erujo6ID9sbGx8nq9ds25Iaitva3tUjV+v1+fffaZjh8/rrNnz160pra29qLjPnXqlE6dOmV/9vv9X2K2AACgswr6FaGjR49q1qxZWrNmjSIjI4PdfbvKz8+X0+m0t7i4uI4eEgAAaEdBD0JVVVVqaGjQDTfcoLCwMIWFhWnbtm1avny5wsLCFBsbq9OnT6upqSnguPr6erlcLkmSy+W64Cmyts+Xq3E4HIqKilLfvn3VrVu3i9a09XG+BQsWyOfz2dvRo0e/8joAAICvv6AHoQkTJmjfvn2qrq62t8TERGVkZNj/3b17d5WVldnH1NXV6ciRI0pOTpYkJScna9++fQFPd5WWlsrhcGjYsGF2zbl9tNW09REeHq5Ro0YF1LS2tqqsrMyuOV9ERIQcDkfABgAAuq6g3yN07bXXavjw4QH7evbsqT59+tj7MzMzlZubq969e8vhcOj+++9XcnKyxowZI0maOHGihg0bpmnTpqmgoEBer1eLFi1SVlaWIiIiJEn33nuvnnvuOc2dO1f33HOPtm7dqhdffFGbNv3fk1i5ubmaPn26EhMTNXr0aC1dulTNzc26++67gz1tAADQCbXLzdKX88wzzyg0NFSTJ0/WqVOnlJKSohdeeMFu79atm4qLi3XfffcpOTlZPXv21PTp0/WrX/3KromPj9emTZv04IMPatmyZbruuuv0hz/8QSkpKXbNlClT9PHHHysvL09er1cjR45USUnJBTdQAwAAM7Xbe4S6At4j1L54jxAAoD18Ld4jBAAA8HXXIX8aAySuiElcFQOAjsYVIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGCnoQys/P14033qhrr71WMTExSktLU11dXUDNyZMnlZWVpT59+uiaa67R5MmTVV9fH1Bz5MgRpaamqkePHoqJidGcOXN05syZgJo33nhDN9xwgyIiIvTtb39bq1atumA8zz//vAYOHKjIyEglJSVpx44dwZ4yAADopIIehLZt26asrCy98847Ki0tVUtLiyZOnKjm5ma75sEHH9Qrr7yiDRs2aNu2bTp27JjuuOMOu/3s2bNKTU3V6dOntX37dq1evVqrVq1SXl6eXXPo0CGlpqbqlltuUXV1tXJycvSLX/xCW7ZssWvWr1+v3NxcPfroo9q9e7dGjBihlJQUNTQ0BHvaAACgEwqxLMtqzxN8/PHHiomJ0bZt2zR+/Hj5fD594xvf0Nq1a/XjH/9YklRbW6uhQ4eqoqJCY8aM0auvvqrbbrtNx44dU2xsrCSpsLBQ8+bN08cff6zw8HDNmzdPmzZtUk1NjX2u9PR0NTU1qaSkRJKUlJSkG2+8Uc8995wkqbW1VXFxcbr//vs1f/78y47d7/fL6XTK5/PJ4XAEe2k0cP6moPeJzuWDJakdPQQA6HKu5Pu73e8R8vl8kqTevXtLkqqqqtTS0iKPx2PXDBkyRP3791dFRYUkqaKiQgkJCXYIkqSUlBT5/X7t37/frjm3j7aatj5Onz6tqqqqgJrQ0FB5PB675nynTp2S3+8P2AAAQNfVrkGotbVVOTk5uummmzR8+HBJktfrVXh4uKKjowNqY2Nj5fV67ZpzQ1Bbe1vbpWr8fr8+++wz/etf/9LZs2cvWtPWx/ny8/PldDrtLS4u7qtNHAAAdArtGoSysrJUU1OjdevWtedpgmbBggXy+Xz2dvTo0Y4eEgAAaEdh7dVxdna2iouLVV5eruuuu87e73K5dPr0aTU1NQVcFaqvr5fL5bJrzn+6q+2psnNrzn/SrL6+Xg6HQ1FRUerWrZu6det20Zq2Ps4XERGhiIiIrzZhAADQ6QT9ipBlWcrOztbGjRu1detWxcfHB7SPGjVK3bt3V1lZmb2vrq5OR44cUXJysiQpOTlZ+/btC3i6q7S0VA6HQ8OGDbNrzu2jraatj/DwcI0aNSqgprW1VWVlZXYNAAAwW9CvCGVlZWnt2rX661//qmuvvda+H8fpdCoqKkpOp1OZmZnKzc1V79695XA4dP/99ys5OVljxoyRJE2cOFHDhg3TtGnTVFBQIK/Xq0WLFikrK8u+YnPvvffqueee09y5c3XPPfdo69atevHFF7Vp0/89iZWbm6vp06crMTFRo0eP1tKlS9Xc3Ky777472NMGAACdUNCD0IoVKyRJ3//+9wP2/+lPf9LPf/5zSdIzzzyj0NBQTZ48WadOnVJKSopeeOEFu7Zbt24qLi7Wfffdp+TkZPXs2VPTp0/Xr371K7smPj5emzZt0oMPPqhly5bpuuuu0x/+8AelpKTYNVOmTNHHH3+svLw8eb1ejRw5UiUlJRfcQA0AAMzU7u8R6sx4jxDaG+8RAoDg+1q9RwgAAODriiAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAwFkEIAAAYiyAEAACMRRACAADGIggBAABjEYQAAICxCEIAAMBYBCEAAGAsghAAADAWQQgAABiLIAQAAIxFEAIAAMYiCAEAAGMRhAAAgLEIQgAAwFgEIQAAYKywjh7A1fD888/rqaeektfr1YgRI/Tss89q9OjRHT0sQAPnb+roIXSoD5akdvQQABiuy18RWr9+vXJzc/Xoo49q9+7dGjFihFJSUtTQ0NDRQwMAAB2syweh3/72t5oxY4buvvtuDRs2TIWFherRo4f++Mc/dvTQAABAB+vSfxo7ffq0qqqqtGDBAntfaGioPB6PKioqLqg/deqUTp06ZX/2+XySJL/f3y7jaz31abv0C3QW7fWzBcBsbb9bLMu6bG2XDkL/+te/dPbsWcXGxgbsj42NVW1t7QX1+fn5euyxxy7YHxcX125jBEzmXNrRIwDQlX3yySdyOp2XrOnSQehKLViwQLm5ufbn1tZWNTY2qk+fPgoJCQnqufx+v+Li4nT06FE5HI6g9t1ZsAasgcQaSKyBxBpIrIEUvDWwLEuffPKJ3G73ZWu7dBDq27evunXrpvr6+oD99fX1crlcF9RHREQoIiIiYF90dHR7DlEOh8PY/+HbsAasgcQaSKyBxBpIrIEUnDW43JWgNl36Zunw8HCNGjVKZWVl9r7W1laVlZUpOTm5A0cGAAC+Drr0FSFJys3N1fTp05WYmKjRo0dr6dKlam5u1t13393RQwMAAB2sywehKVOm6OOPP1ZeXp68Xq9GjhypkpKSC26gvtoiIiL06KOPXvCnOJOwBqyBxBpIrIHEGkisgdQxaxBifZlnywAAALqgLn2PEAAAwKUQhAAAgLEIQgAAwFgEIQAAYCyCUAd4/vnnNXDgQEVGRiopKUk7duzo6CEFTX5+vm688UZde+21iomJUVpamurq6gJqTp48qaysLPXp00fXXHONJk+efMFLL48cOaLU1FT16NFDMTExmjNnjs6cOXM1pxIUS5YsUUhIiHJycux9psz/o48+0s9+9jP16dNHUVFRSkhI0K5du+x2y7KUl5enfv36KSoqSh6PRwcPHgzoo7GxURkZGXI4HIqOjlZmZqZOnDhxtafylZw9e1aPPPKI4uPjFRUVpW9961v69a9/HfBvH3W1NSgvL9ftt98ut9utkJAQFRUVBbQHa7579+7VzTffrMjISMXFxamgoKC9p/alXWoNWlpaNG/ePCUkJKhnz55yu9266667dOzYsYA+uvIanO/ee+9VSEiIli5dGrD/qq6Bhatq3bp1Vnh4uPXHP/7R2r9/vzVjxgwrOjraqq+v7+ihBUVKSor1pz/9yaqpqbGqq6utSZMmWf3797dOnDhh19x7771WXFycVVZWZu3atcsaM2aMNXbsWLv9zJkz1vDhwy2Px2Pt2bPH2rx5s9W3b19rwYIFHTGlr2zHjh3WwIEDre9+97vWrFmz7P0mzL+xsdEaMGCA9fOf/9yqrKy0/vGPf1hbtmyx/v73v9s1S5YssZxOp1VUVGS9++671g9/+EMrPj7e+uyzz+yaW2+91RoxYoT1zjvvWG+++ab17W9/25o6dWpHTOmKPfHEE1afPn2s4uJi69ChQ9aGDRusa665xlq2bJld09XWYPPmzdbChQutl156yZJkbdy4MaA9GPP1+XxWbGyslZGRYdXU1Fh/+ctfrKioKOt3v/vd1ZrmJV1qDZqamiyPx2OtX7/eqq2ttSoqKqzRo0dbo0aNCuijK6/BuV566SVrxIgRltvttp555pmAtqu5BgShq2z06NFWVlaW/fns2bOW2+228vPzO3BU7aehocGSZG3bts2yrM9/EXTv3t3asGGDXXPgwAFLklVRUWFZ1uc/RKGhoZbX67VrVqxYYTkcDuvUqVNXdwJf0SeffGINGjTIKi0ttb73ve/ZQciU+c+bN88aN27cF7a3trZaLpfLeuqpp+x9TU1NVkREhPWXv/zFsizLeu+99yxJ1s6dO+2aV1991QoJCbE++uij9ht8kKSmplr33HNPwL477rjDysjIsCyr66/B+V+AwZrvCy+8YPXq1SvgZ2HevHnW4MGD23lGV+5SIaDNjh07LEnW4cOHLcsyZw0+/PBD65vf/KZVU1NjDRgwICAIXe014E9jV9Hp06dVVVUlj8dj7wsNDZXH41FFRUUHjqz9+Hw+SVLv3r0lSVVVVWppaQlYgyFDhqh///72GlRUVCghISHgpZcpKSny+/3av3//VRz9V5eVlaXU1NSAeUrmzP/ll19WYmKifvKTnygmJkbXX3+9fv/739vthw4dktfrDVgHp9OppKSkgHWIjo5WYmKiXePxeBQaGqrKysqrN5mvaOzYsSorK9P7778vSXr33Xf11ltv6Qc/+IEkM9bgXMGab0VFhcaPH6/w8HC7JiUlRXV1dTp+/PhVmk3w+Hw+hYSE2P+upQlr0NraqmnTpmnOnDn6zne+c0H71V4DgtBV9K9//Utnz5694K3WsbGx8nq9HTSq9tPa2qqcnBzddNNNGj58uCTJ6/UqPDz8gn/M9tw18Hq9F12jtravu3Xr1mn37t3Kz8+/oM2E+UvSP/7xD61YsUKDBg3Sli1bdN999+mBBx7Q6tWrJf3fPC71s+D1ehUTExPQHhYWpt69e3eKdZg/f77S09M1ZMgQde/eXddff71ycnKUkZEhyYw1OFew5tsVfj7anDx5UvPmzdPUqVPtf2DUhDV48sknFRYWpgceeOCi7Vd7Dbr8P7GBjpOVlaWamhq99dZbHT2Uq+bo0aOaNWuWSktLFRkZ2dHD6TCtra1KTEzUb37zG0nS9ddfr5qaGhUWFmr69OkdPLqr48UXX9SaNWu0du1afec731F1dbVycnLkdruNWQN8sZaWFv30pz+VZVlasWJFRw/nqqmqqtKyZcu0e/duhYSEdPRwJHFF6Krq27evunXrdsETQvX19XK5XB00qvaRnZ2t4uJivf7667ruuuvs/S6XS6dPn1ZTU1NA/blr4HK5LrpGbW1fZ1VVVWpoaNANN9ygsLAwhYWFadu2bVq+fLnCwsIUGxvbpeffpl+/fho2bFjAvqFDh+rIkSOS/m8el/pZcLlcamhoCGg/c+aMGhsbO8U6zJkzx74qlJCQoGnTpunBBx+0rxSasAbnCtZ8u8LPR1sIOnz4sEpLS+2rQVLXX4M333xTDQ0N6t+/v/078vDhw5o9e7YGDhwo6eqvAUHoKgoPD9eoUaNUVlZm72ttbVVZWZmSk5M7cGTBY1mWsrOztXHjRm3dulXx8fEB7aNGjVL37t0D1qCurk5Hjhyx1yA5OVn79u0L+EFo+2Vx/pfr182ECRO0b98+VVdX21tiYqIyMjLs/+7K829z0003XfDahPfff18DBgyQJMXHx8vlcgWsg9/vV2VlZcA6NDU1qaqqyq7ZunWrWltblZSUdBVm8e/59NNPFRoa+Cu2W7duam1tlWTGGpwrWPNNTk5WeXm5Wlpa7JrS0lINHjxYvXr1ukqz+eraQtDBgwf1t7/9TX369Alo7+prMG3aNO3duzfgd6Tb7dacOXO0ZcsWSR2wBld8ezX+LevWrbMiIiKsVatWWe+99541c+ZMKzo6OuAJoc7svvvus5xOp/XGG29Y//znP+3t008/tWvuvfdeq3///tbWrVutXbt2WcnJyVZycrLd3vb4+MSJE63q6mqrpKTE+sY3vtGpHh8/17lPjVmWGfPfsWOHFRYWZj3xxBPWwYMHrTVr1lg9evSw/vznP9s1S5YssaKjo62//vWv1t69e63/+q//uuij1Ndff71VWVlpvfXWW9agQYO+to+On2/69OnWN7/5Tfvx+Zdeesnq27evNXfuXLumq63BJ598Yu3Zs8fas2ePJcn67W9/a+3Zs8d+IioY821qarJiY2OtadOmWTU1Nda6deusHj16fG0eHb/UGpw+fdr64Q9/aF133XVWdXV1wO/Ic59+6sprcDHnPzVmWVd3DQhCHeDZZ5+1+vfvb4WHh1ujR4+23nnnnY4eUtBIuuj2pz/9ya757LPPrF/+8pdWr169rB49elg/+tGPrH/+858B/XzwwQfWD37wAysqKsrq27evNXv2bKulpeUqzyY4zg9Cpsz/lVdesYYPH25FRERYQ4YMsVauXBnQ3traaj3yyCNWbGysFRERYU2YMMGqq6sLqPl//+//WVOnTrWuueYay+FwWHfffbf1ySefXM1pfGV+v9+aNWuW1b9/fysyMtL6j//4D2vhwoUBX3hdbQ1ef/31i/78T58+3bKs4M333XfftcaNG2dFRERY3/zmN60lS5ZcrSle1qXW4NChQ1/4O/L111+3++jKa3AxFwtCV3MNQizrnNecAgAAGIR7hAAAgLEIQgAAwFgEIQAAYCyCEAAAMBZBCAAAGIsgBAAAjEUQAgAAxiIIAQAAYxGEAACAsQhCAADAWAQhAABgLIIQAAAw1v8HDbjRMJahj4MAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# find what character length covers 95% of sequences\n",
        "output_seq_char_len = int(np.percentile(char_lens, 95))\n",
        "output_seq_char_len"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V7DjUfkCPUr4",
        "outputId": "eed3657d-5599-4a23-ec48-2bc7a670dca8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "290"
            ]
          },
          "metadata": {},
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# get all keyboard characters\n",
        "import string\n",
        "alphabet = string.ascii_lowercase + string.digits + string.punctuation\n",
        "alphabet"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "WM9OKe-bPUoS",
        "outputId": "86052cc4-69d5-462d-b7b1-c964a1d10e7f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'abcdefghijklmnopqrstuvwxyz0123456789!\"#$%&\\'()*+,-./:;<=>?@[\\\\]^_`{|}~'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 62
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# creating char-level token vectorizer instance\n",
        "NUM_CHAR_TOKENS = len(alphabet) + 2 # add 2 for space\n",
        "\n",
        "char_vectorizer = TextVectorization(max_tokens=NUM_CHAR_TOKENS,\n",
        "                                    output_sequence_length=output_seq_char_len,\n",
        "                                    standardize=\"lower_and_strip_punctuation\",\n",
        "                                    # standardize=None, # set none if we want to include punctuation\n",
        "                                    name=\"char_vectorizer\")"
      ],
      "metadata": {
        "id": "9a4cjhM7PUkj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# adapt character vectorizer to training character\n",
        "char_vectorizer.adapt(train_chars)"
      ],
      "metadata": {
        "id": "gMLfc4o2PUex"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# check vocab stats\n",
        "char_vocab = char_vectorizer.get_vocabulary()\n",
        "print(f\"Number of different characters in character vocab: {len(char_vocab)}\")\n",
        "print(f\"5 most common characters: {char_vocab[:5]}\")\n",
        "print(f\"5 least common characters: {char_vocab[-5:]}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xR5-Dl-aPUbE",
        "outputId": "337b4db7-a3af-4299-a146-94f6c32bba0a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of different characters in character vocab: 28\n",
            "5 most common characters: ['', '[UNK]', 'e', 't', 'i']\n",
            "5 least common characters: ['k', 'x', 'z', 'q', 'j']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Test out character vectorizer\n",
        "random_train_chars = random.choice(train_chars)\n",
        "print(f\"Charified text:\\n{random_train_chars}\")\n",
        "print(f\"\\nLength of chars: {len(random_train_chars.split())}\")\n",
        "vectorized_chars = char_vectorizer([random_train_chars])\n",
        "print(f\"\\nVectorized chars:\\n{vectorized_chars}\")\n",
        "print(f\"\\nLength of vectorized chars: {len(vectorized_chars[0])}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GTD-E5UfBIjG",
        "outputId": "a650aebf-a16b-4d16-b539-dc91cd5fa67e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Charified text:\n",
            "i n   t h i s   m u l t i c e n t r e   ,   r a n d o m i s e d   ,   d o u b l e - m a s k e d   ,   p l a c e b o - c o n t r o l l e d   t r i a l   ,   w e   e n r o l l e d   c o r t i c o s t e r o i d   u n r e s p o n s i v e   a d u l t   p a t i e n t s   (   a g e d   @   y e a r s   )   w i t h   p r i m a r y   i m m u n e   t h r o m b o c y t o p e n i a   a n d   a   p l a t e l e t   c o u n t   o f   l e s s   t h a n   @   (   @   )   p l a t e l e t s   p e r   l .   p a t i e n t s   w e r e   r a n d o m l y   a s s i g n e d   (   @ : @   )   t o   f o u r   w e e k l y   i n f u s i o n s   o f   @   m g / m   (   @   )   r i t u x i m a b   o r   p l a c e b o   .\n",
            "\n",
            "Length of chars: 289\n",
            "\n",
            "Vectorized chars:\n",
            "[[ 4  6  3 13  4  9 15 16 12  3  4 11  2  6  3  8  2  8  5  6 10  7 15  4\n",
            "   9  2 10 10  7 16 22 12  2 15  5  9 23  2 10 14 12  5 11  2 22  7 11  7\n",
            "   6  3  8  7 12 12  2 10  3  8  4  5 12 20  2  2  6  8  7 12 12  2 10 11\n",
            "   7  8  3  4 11  7  9  3  2  8  7  4 10 16  6  8  2  9 14  7  6  9  4 21\n",
            "   2  5 10 16 12  3 14  5  3  4  2  6  3  9  5 18  2 10 19  2  5  8  9 20\n",
            "   4  3 13 14  8  4 15  5  8 19  4 15 15 16  6  2  3 13  8  7 15 22  7 11\n",
            "  19  3  7 14  2  6  4  5  5  6 10  5 14 12  5  3  2 12  2  3 11  7 16  6\n",
            "   3  7 17 12  2  9  9  3 13  5  6 14 12  5  3  2 12  2  3  9 14  2  8 12\n",
            "  14  5  3  4  2  6  3  9 20  2  8  2  8  5  6 10  7 15 12 19  5  9  9  4\n",
            "  18  6  2 10  3  7 17  7 16  8 20  2  2 23 12 19  4  6 17 16  9  4  7  6\n",
            "   9  7 17 15 18 15  8  4  3 16 24  4 15  5 22  7  8 14 12  5 11  2 22  7\n",
            "   0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0\n",
            "   0  0]]\n",
            "\n",
            "Length of vectorized chars: 290\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating an character-level embedding layer"
      ],
      "metadata": {
        "id": "KA23IRe8DRGP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "char_embed = tf.keras.layers.Embedding(input_dim= NUM_CHAR_TOKENS,\n",
        "                                       output_dim= 25,\n",
        "                                       mask_zero=False,\n",
        "                                       name=\"char_embed\")\n",
        "\n",
        "print(f\"Charified text (before vectorization and embedding):\\n{random_train_chars}\\n\")\n",
        "char_embed_example = char_embed(char_vectorizer([random_train_chars]))\n",
        "print(f\"Embedded chars (after vectorization and embedding):\\n{char_embed_example}\\n\")\n",
        "print(f\"Character embedding shape: {char_embed_example.shape}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P2JCrNxFBIf0",
        "outputId": "1e6a4eaf-ce20-46ec-8aa0-db9952b4d3ef"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Charified text (before vectorization and embedding):\n",
            "i n   t h i s   m u l t i c e n t r e   ,   r a n d o m i s e d   ,   d o u b l e - m a s k e d   ,   p l a c e b o - c o n t r o l l e d   t r i a l   ,   w e   e n r o l l e d   c o r t i c o s t e r o i d   u n r e s p o n s i v e   a d u l t   p a t i e n t s   (   a g e d   @   y e a r s   )   w i t h   p r i m a r y   i m m u n e   t h r o m b o c y t o p e n i a   a n d   a   p l a t e l e t   c o u n t   o f   l e s s   t h a n   @   (   @   )   p l a t e l e t s   p e r   l .   p a t i e n t s   w e r e   r a n d o m l y   a s s i g n e d   (   @ : @   )   t o   f o u r   w e e k l y   i n f u s i o n s   o f   @   m g / m   (   @   )   r i t u x i m a b   o r   p l a c e b o   .\n",
            "\n",
            "Embedded chars (after vectorization and embedding):\n",
            "[[[ 0.01734987  0.04088182  0.03298097 ... -0.03720428 -0.02333094\n",
            "   -0.0351436 ]\n",
            "  [-0.01253979 -0.00867556  0.04898559 ...  0.00443391  0.04133997\n",
            "   -0.02626204]\n",
            "  [ 0.01277477  0.01247247  0.03468876 ...  0.00890012  0.02771593\n",
            "   -0.03062971]\n",
            "  ...\n",
            "  [-0.02073165  0.03056205  0.04664818 ... -0.02693449  0.00909891\n",
            "    0.00852178]\n",
            "  [-0.02073165  0.03056205  0.04664818 ... -0.02693449  0.00909891\n",
            "    0.00852178]\n",
            "  [-0.02073165  0.03056205  0.04664818 ... -0.02693449  0.00909891\n",
            "    0.00852178]]]\n",
            "\n",
            "Character embedding shape: (1, 290, 25)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model 3 (Conv1D)"
      ],
      "metadata": {
        "id": "bHKxQjp5jT9l"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# make Conv1D on chars only\n",
        "inputs = tf.keras.layers.Input(shape=(1,), dtype=tf.string)\n",
        "char_vectors = char_vectorizer(inputs)\n",
        "char_embeddings = char_embed(char_vectors)\n",
        "x = tf.keras.layers.Conv1D(64, kernel_size=5, padding=\"same\", activation=\"relu\")(char_embeddings)\n",
        "x = tf.keras.layers.GlobalAveragePooling1D()(x)\n",
        "outputs = tf.keras.layers.Dense(num_classes, activation=\"softmax\")(x)\n",
        "\n",
        "model_3 = tf.keras.Model(inputs=inputs,\n",
        "                         outputs=outputs,\n",
        "                         name=\"model_3_conv1d_char_embeddings\")\n",
        "\n",
        "# compile\n",
        "model_3.compile(loss=\"categorical_crossentropy\",\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])\n"
      ],
      "metadata": {
        "id": "5xj1Vd-ZBIcT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_3.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mFb_1HcnBIYz",
        "outputId": "872e1f4f-fb6a-4ebd-86a7-45f5cfc0e1b3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3_conv1d_char_embeddings\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_3 (InputLayer)        [(None, 1)]               0         \n",
            "                                                                 \n",
            " char_vectorizer (TextVecto  (None, 290)               0         \n",
            " rization)                                                       \n",
            "                                                                 \n",
            " char_embed (Embedding)      (None, 290, 25)           1750      \n",
            "                                                                 \n",
            " conv1d_1 (Conv1D)           (None, 290, 64)           8064      \n",
            "                                                                 \n",
            " global_average_pooling1d_1  (None, 64)                0         \n",
            "  (GlobalAveragePooling1D)                                       \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 5)                 325       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 10139 (39.61 KB)\n",
            "Trainable params: 10139 (39.61 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# creating char_level_datasets\n",
        "train_char_dataset = tf.data.Dataset.from_tensor_slices((train_chars, train_labels_one_hot)).batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "val_char_dataset = tf.data.Dataset.from_tensor_slices((val_chars, val_labels_one_hot)).batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "test_char_dataset = tf.data.Dataset.from_tensor_slices((test_chars, test_labels_one_hot)).batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "\n",
        "train_char_dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uBtxNerCBIVP",
        "outputId": "c72ba6d5-d4d8-442e-d232-16800bf360f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<_PrefetchDataset element_spec=(TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Fit the model\n",
        "model_3_history = model_3.fit(train_char_dataset,\n",
        "                              steps_per_epoch=int(0.1*len(train_char_dataset)),\n",
        "                              epochs=3,\n",
        "                              validation_data=val_char_dataset,\n",
        "                              validation_steps=int(0.1 * len(val_char_dataset)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mYkVantpBIRp",
        "outputId": "529a44d7-0425-4cd1-ee69-090db9fbfc1c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "562/562 [==============================] - 6s 8ms/step - loss: 1.4605 - accuracy: 0.3453 - val_loss: 1.4105 - val_accuracy: 0.3976\n",
            "Epoch 2/3\n",
            "562/562 [==============================] - 3s 5ms/step - loss: 1.3742 - accuracy: 0.4141 - val_loss: 1.3427 - val_accuracy: 0.4405\n",
            "Epoch 3/3\n",
            "562/562 [==============================] - 3s 5ms/step - loss: 1.3332 - accuracy: 0.4480 - val_loss: 1.3163 - val_accuracy: 0.4471\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_3.evaluate(val_char_dataset)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "udBc3i45BIOS",
        "outputId": "38f543d1-b54e-49de-8fd7-5fa92a0f9a8f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 3s 3ms/step - loss: 1.3161 - accuracy: 0.4481\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.3160841464996338, 0.4481332004070282]"
            ]
          },
          "metadata": {},
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_3_pred_probs = model_3.predict(val_char_dataset)\n",
        "model_3_pred_probs"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t0rJT9KwPUW_",
        "outputId": "3e464ed5-b85b-42ce-abdb-5c8050602781"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 2s 3ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.1676105 , 0.2496262 , 0.25604883, 0.18530549, 0.14140901],\n",
              "       [0.11392321, 0.16502018, 0.16016495, 0.06585773, 0.49503392],\n",
              "       [0.10147274, 0.16171625, 0.41223532, 0.10716747, 0.21740822],\n",
              "       ...,\n",
              "       [0.07665232, 0.11550133, 0.24905907, 0.04734882, 0.5114385 ],\n",
              "       [0.05768117, 0.10047982, 0.2412954 , 0.0444183 , 0.5561253 ],\n",
              "       [0.142175  , 0.17853574, 0.21677431, 0.07356828, 0.38894662]],\n",
              "      dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_3_preds = tf.argmax(model_3_pred_probs, axis=1)\n",
        "model_3_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wDWJlnxFPUTW",
        "outputId": "cde3e4dd-f76b-43de-e667-f11702670f30"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(30212,), dtype=int64, numpy=array([2, 4, 2, ..., 4, 4, 4])>"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_3_results = calculate_results(val_labels_encoded,\n",
        "                                    model_3_preds)\n",
        "model_3_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "slrrztT9PUPq",
        "outputId": "8c4eaeb9-7798-4fe0-d63d-a317504c2724"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1344: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 44.813319210909576,\n",
              " 'precision': 0.3765038613993536,\n",
              " 'recall': 0.44813319210909575,\n",
              " 'f1': 0.3942968004834402}"
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Model 4: Combining pretrained token embeddings + character embeddings (hybrid embedding layer)\n",
        "\n",
        "\n",
        "1.   Create a token-level model (similar to model_1)\n",
        "2.   Create a character-level model (similar to model_3 with a slight modification to reflect the paper)\n",
        "3.   Combine (using layers.Concatenate) the outputs of 1 and 2\n",
        "4.   Build a series of output layers on top of 3 similar to Figure 1 and section 4.2 of *Neural Networks for Joint Sentence Classification*\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "VOgylhrYDeEe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 1. Creating token inputs/model\n",
        "token_inputs = tf.keras.layers.Input(shape=[], dtype=tf.string, name=\"token_string\")\n",
        "token_embeddings = tf_hub_embedding_layer(token_inputs)\n",
        "token_output = tf.keras.layers.Dense(128, activation=\"relu\")(token_embeddings)\n",
        "token_model = tf.keras.Model(inputs=token_inputs,\n",
        "                             outputs=token_output)\n",
        "\n",
        "# 2. Creating character inputs/model\n",
        "char_inputs = tf.keras.layers.Input(shape=(1,), dtype=tf.string, name=\"char_input\")\n",
        "char_vectors = char_vectorizer(char_inputs)\n",
        "char_embeddings = char_embed(char_vectors)\n",
        "char_bi_lstm = tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(25))(char_embeddings) # bi-LSTM shown in Figure 1 of https://arxiv.org/pdf/1612.05251.pdf\n",
        "char_model = tf.keras.Model(inputs=char_inputs,\n",
        "                            outputs=char_bi_lstm)\n",
        "\n",
        "# 3. Combining the layers using layers.Concatenate\n",
        "token_char_concat = tf.keras.layers.Concatenate(name=\"token_char_hybrid\")([token_model.output,\n",
        "                                                                           char_model.output])\n",
        "\n",
        "# 4. Create output layer\n",
        "combined_dropout = tf.keras.layers.Dropout(0.5)(token_char_concat)\n",
        "combined_dense = tf.keras.layers.Dense(200, activation=\"relu\")(combined_dropout)\n",
        "final_dropout = tf.keras.layers.Dropout(0.5)(combined_dense)\n",
        "output_layer = tf.keras.layers.Dense(num_classes, activation=\"softmax\")(final_dropout)\n",
        "\n",
        "# 5. Construct the model\n",
        "model_4 = tf.keras.Model(inputs=[token_model.input, char_model.input],\n",
        "                         outputs=output_layer,\n",
        "                         name=\"model_4_token_and_char_embeddings\")\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "vZ9SBI7TPULf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_4.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PQDBdz-RPUHl",
        "outputId": "e3e61980-b41d-4d82-a334-56891353722b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_4_token_and_char_embeddings\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                Output Shape                 Param #   Connected to                  \n",
            "==================================================================================================\n",
            " char_input (InputLayer)     [(None, 1)]                  0         []                            \n",
            "                                                                                                  \n",
            " token_string (InputLayer)   [(None,)]                    0         []                            \n",
            "                                                                                                  \n",
            " char_vectorizer (TextVecto  (None, 290)                  0         ['char_input[0][0]']          \n",
            " rization)                                                                                        \n",
            "                                                                                                  \n",
            " universal_sentence_encoder  (None, 512)                  2567978   ['token_string[0][0]']        \n",
            "  (KerasLayer)                                            24                                      \n",
            "                                                                                                  \n",
            " char_embed (Embedding)      (None, 290, 25)              1750      ['char_vectorizer[1][0]']     \n",
            "                                                                                                  \n",
            " dense_4 (Dense)             (None, 128)                  65664     ['universal_sentence_encoder[1\n",
            "                                                                    ][0]']                        \n",
            "                                                                                                  \n",
            " bidirectional (Bidirection  (None, 50)                   10200     ['char_embed[1][0]']          \n",
            " al)                                                                                              \n",
            "                                                                                                  \n",
            " token_char_hybrid (Concate  (None, 178)                  0         ['dense_4[0][0]',             \n",
            " nate)                                                               'bidirectional[0][0]']       \n",
            "                                                                                                  \n",
            " dropout (Dropout)           (None, 178)                  0         ['token_char_hybrid[0][0]']   \n",
            "                                                                                                  \n",
            " dense_5 (Dense)             (None, 200)                  35800     ['dropout[0][0]']             \n",
            "                                                                                                  \n",
            " dropout_1 (Dropout)         (None, 200)                  0         ['dense_5[0][0]']             \n",
            "                                                                                                  \n",
            " dense_6 (Dense)             (None, 5)                    1005      ['dropout_1[0][0]']           \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 256912243 (980.04 MB)\n",
            "Trainable params: 114419 (446.95 KB)\n",
            "Non-trainable params: 256797824 (979.61 MB)\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(model_4)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 856
        },
        "id": "NlVVVSULPUDz",
        "outputId": "9842cd75-c319-434b-bf4f-976fd8fc4ac3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_4.compile(loss=\"categorical_crossentropy\",\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "zMU3kSK8PUAO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Combining token and character data into a tf.data dataset\n",
        "The difference with this model is that it requires two inputs, token-level sequences and character-level sequences.\n",
        "\n",
        "We can do this by create a tf.data.Dataset with a tuple as it's first input, for example:\n",
        "\n",
        "((token_data, char_data), (label))"
      ],
      "metadata": {
        "id": "dO34sU_E2Vwk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_char_token_data = tf.data.Dataset.from_tensor_slices((train_sentences,train_chars))\n",
        "train_char_token_labels = tf.data.Dataset.from_tensor_slices(train_labels_one_hot)\n",
        "train_char_token_dataset = tf.data.Dataset.zip((train_char_token_data, train_char_token_labels))\n",
        "\n",
        "train_char_token_dataset = train_char_token_dataset.batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "\n",
        "val_char_token_data = tf.data.Dataset.from_tensor_slices((val_sentences, val_chars))\n",
        "val_char_token_labels = tf.data.Dataset.from_tensor_slices(val_labels_one_hot)\n",
        "val_char_token_dataset = tf.data.Dataset.zip((val_char_token_data, val_char_token_labels))\n",
        "val_char_token_dataset = val_char_token_dataset.batch(32).prefetch(tf.data.AUTOTUNE)"
      ],
      "metadata": {
        "id": "bz_ojsgmPT8b"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_char_token_dataset, val_char_token_dataset\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zEXQI2UAPT4Y",
        "outputId": "00701c57-9856-4143-a48f-e525a804b7b3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<_PrefetchDataset element_spec=((TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None)), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>,\n",
              " <_PrefetchDataset element_spec=((TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None)), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>)"
            ]
          },
          "metadata": {},
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_4_history = model_4.fit(train_char_token_dataset,\n",
        "                              steps_per_epoch=int(0.1*len(train_char_token_dataset)),\n",
        "                              epochs=3,\n",
        "                              validation_data=val_char_token_dataset,\n",
        "                               validation_steps=int(0.1 * len(val_char_token_dataset)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TCqMmmW77m_w",
        "outputId": "e15526e2-f7d5-40eb-9986-fc3fa4c2a1a1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "562/562 [==============================] - 30s 42ms/step - loss: 0.9719 - accuracy: 0.6115 - val_loss: 0.8049 - val_accuracy: 0.6935\n",
            "Epoch 2/3\n",
            "562/562 [==============================] - 19s 35ms/step - loss: 0.8060 - accuracy: 0.6888 - val_loss: 0.7251 - val_accuracy: 0.7267\n",
            "Epoch 3/3\n",
            "562/562 [==============================] - 23s 40ms/step - loss: 0.7769 - accuracy: 0.7021 - val_loss: 0.6933 - val_accuracy: 0.7384\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_4.evaluate(val_char_token_dataset)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ME7ALfAW7m6V",
        "outputId": "5287f130-aa7a-4c11-99f9-77e97eb82b6a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 18s 19ms/step - loss: 0.6980 - accuracy: 0.7308\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.6979666352272034, 0.7308354377746582]"
            ]
          },
          "metadata": {},
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_4_pred_probs = model_4.predict(val_char_token_dataset)\n",
        "model_4_preds = tf.argmax(model_4_pred_probs, axis=1)\n",
        "model_4_results = calculate_results(y_true=val_labels_encoded,\n",
        "                                    y_pred=model_4_preds)\n",
        "model_4_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lQL75w777m0w",
        "outputId": "dfa30b82-c0b8-44cb-d2c5-994e90bf79cc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 20s 20ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 73.08354296306103,\n",
              " 'precision': 0.7289075066892654,\n",
              " 'recall': 0.7308354296306103,\n",
              " 'f1': 0.7276069458682359}"
            ]
          },
          "metadata": {},
          "execution_count": 84
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Positional embeddings"
      ],
      "metadata": {
        "id": "2ToKGXQuIzFD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_df.head()"
      ],
      "metadata": {
        "id": "keySVqc07mvk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "332d62af-5bc3-482c-9b92-138a0a0cd474"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      target                                               text  line_number  \\\n",
              "0  OBJECTIVE  to investigate the efficacy of @ weeks of dail...            0   \n",
              "1    METHODS  a total of @ patients with primary knee oa wer...            1   \n",
              "2    METHODS  outcome measures included pain reduction and i...            2   \n",
              "3    METHODS  pain was assessed using the visual analog pain...            3   \n",
              "4    METHODS  secondary outcome measures included the wester...            4   \n",
              "\n",
              "   total_lines  \n",
              "0           11  \n",
              "1           11  \n",
              "2           11  \n",
              "3           11  \n",
              "4           11  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1a667798-66e6-4de7-86ae-b1d493627294\" 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>target</th>\n",
              "      <th>text</th>\n",
              "      <th>line_number</th>\n",
              "      <th>total_lines</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>OBJECTIVE</td>\n",
              "      <td>to investigate the efficacy of @ weeks of dail...</td>\n",
              "      <td>0</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>a total of @ patients with primary knee oa wer...</td>\n",
              "      <td>1</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>outcome measures included pain reduction and i...</td>\n",
              "      <td>2</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>pain was assessed using the visual analog pain...</td>\n",
              "      <td>3</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>secondary outcome measures included the wester...</td>\n",
              "      <td>4</td>\n",
              "      <td>11</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-1a667798-66e6-4de7-86ae-b1d493627294')\"\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-1a667798-66e6-4de7-86ae-b1d493627294 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-1a667798-66e6-4de7-86ae-b1d493627294');\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-c75b3fdf-ca4c-4c4d-b357-4df58ef92b90\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-c75b3fdf-ca4c-4c4d-b357-4df58ef92b90')\"\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-c75b3fdf-ca4c-4c4d-b357-4df58ef92b90 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 85
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Adding the line number feature"
      ],
      "metadata": {
        "id": "vWtvc_dYI1gh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_df[\"line_number\"].value_counts()"
      ],
      "metadata": {
        "id": "ylgvLrcT7mmh",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b747c6d5-ff26-4d9a-9eb3-5f7dc1708e92"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0     15000\n",
              "1     15000\n",
              "2     15000\n",
              "3     15000\n",
              "4     14992\n",
              "5     14949\n",
              "6     14758\n",
              "7     14279\n",
              "8     13346\n",
              "9     11981\n",
              "10    10041\n",
              "11     7892\n",
              "12     5853\n",
              "13     4152\n",
              "14     2835\n",
              "15     1861\n",
              "16     1188\n",
              "17      751\n",
              "18      462\n",
              "19      286\n",
              "20      162\n",
              "21      101\n",
              "22       66\n",
              "23       33\n",
              "24       22\n",
              "25       14\n",
              "26        7\n",
              "27        4\n",
              "28        3\n",
              "29        1\n",
              "30        1\n",
              "Name: line_number, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_df[\"line_number\"].plot.hist()"
      ],
      "metadata": {
        "id": "9k83SdMM7mhS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "outputId": "31916bde-f832-422f-bf8b-524e77472e8b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: ylabel='Frequency'>"
            ]
          },
          "metadata": {},
          "execution_count": 87
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# using tensorflow to create one-hot encoder\n",
        "train_line_numbers_one_hot = tf.one_hot(train_df[\"line_number\"].to_numpy(), depth=15)\n",
        "val_line_numbers_one_hot = tf.one_hot(val_df[\"line_number\"].to_numpy(), depth=15)\n",
        "test_line_numbers_one_hot = tf.one_hot(test_df[\"line_number\"].to_numpy(), depth=15)\n",
        "train_line_numbers_one_hot[:10], train_line_numbers_one_hot.shape\n"
      ],
      "metadata": {
        "id": "Xe_9fYCR7mcC",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d75748ff-1c3b-44cc-8e9f-7d8a318e6ebf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(10, 15), dtype=float32, numpy=\n",
              " array([[1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0.],\n",
              "        [0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.]],\n",
              "       dtype=float32)>,\n",
              " TensorShape([180040, 15]))"
            ]
          },
          "metadata": {},
          "execution_count": 88
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_df[\"total_lines\"].value_counts()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9yp-IP7DGII0",
        "outputId": "31f7b6f9-49bf-4bab-c830-88ce8c6fccc5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "11    24468\n",
              "10    23639\n",
              "12    22113\n",
              "9     19400\n",
              "13    18438\n",
              "14    14610\n",
              "8     12285\n",
              "15    10768\n",
              "7      7464\n",
              "16     7429\n",
              "17     5202\n",
              "6      3353\n",
              "18     3344\n",
              "19     2480\n",
              "20     1281\n",
              "5      1146\n",
              "21      770\n",
              "22      759\n",
              "23      264\n",
              "4       215\n",
              "24      200\n",
              "25      182\n",
              "26       81\n",
              "28       58\n",
              "3        32\n",
              "30       31\n",
              "27       28\n",
              "Name: total_lines, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 89
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_df.total_lines.plot.hist();"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 431
        },
        "id": "ybqMVAZcGIFM",
        "outputId": "abd511be-c9d5-47d8-eab8-d6cc2d21f102"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.percentile(train_df.total_lines, 98)"
      ],
      "metadata": {
        "id": "XnTgoNQB7mWx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3810e2fb-7756-411e-bb2f-3a2b57769700"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "20.0"
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_total_lines_one_hot = tf.one_hot(train_df[\"total_lines\"].to_numpy(), depth=20)\n",
        "val_total_lines_one_hot = tf.one_hot(val_df[\"total_lines\"].to_numpy(), depth=20)\n",
        "test_total_lines_one_hot = tf.one_hot(test_df[\"total_lines\"].to_numpy(), depth=20)"
      ],
      "metadata": {
        "id": "jQPdSjok7mRY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_total_lines_one_hot"
      ],
      "metadata": {
        "id": "ZsA0IhcH7mMA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1ffdb29a-b730-4b37-8db9-71ab2a2eb7d3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(180040, 20), dtype=float32, numpy=\n",
              "array([[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.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Building a tribrid embedding model\n",
        "\n",
        "1. Create a token-level mode\n",
        "2. Create a character-level mode\n",
        "3. Create a model for \"line_number\" feature\n",
        "4. Create a model for the \"total_lines\" feature\n",
        "5. Combine outputs of 1 & 2 using tf.keras.layers.Concatenate\n",
        "6. Combine outputs of 3 & 4 & 5 using tf.keras.layers.Concatenate\n",
        "7. Create an output layer to accept tribried embedding and output label probablities\n",
        "8. Combine the inputs of 1,2,3,4 and outputs of into a tf.keras.Model"
      ],
      "metadata": {
        "id": "vJ2FrM6aV3mv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 1. Token inputs\n",
        "token_inputs = tf.keras.layers.Input(shape=[], dtype=tf.string, name=\"token_inputs\")\n",
        "token_embeddings = tf_hub_embedding_layer(token_inputs)\n",
        "token_outputs = tf.keras.layers.Dense(128, activation=\"relu\")(token_embeddings)\n",
        "token_model = tf.keras.Model(token_inputs, token_outputs)\n",
        "\n",
        "# 2. Character inputs\n",
        "char_inputs = tf.keras.layers.Input(shape=(1,), dtype=tf.string, name=\"char_inputs\")\n",
        "char_vectors = char_vectorizer(char_inputs)\n",
        "char_embeddings = char_embed(char_vectors)\n",
        "char_bi_lstm = tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32))(char_embeddings)\n",
        "char_model = tf.keras.Model(char_inputs, char_bi_lstm)\n",
        "\n",
        "# 3. Line numbers inputs\n",
        "line_number_inputs = tf.keras.layers.Input(shape=(15,), dtype=tf.int32, name=\"line_number_input\")\n",
        "x = tf.keras.layers.Dense(32, activation=\"relu\")(line_number_inputs)\n",
        "line_number_model = tf.keras.Model(inputs=line_number_inputs,\n",
        "                                   outputs=x)\n",
        "\n",
        "# 4. Total lines inputs\n",
        "total_lines_inputs = tf.keras.layers.Input(shape=(20,), dtype=tf.int32, name=\"total_lines_input\")\n",
        "y = tf.keras.layers.Dense(32, activation=\"relu\")(total_lines_inputs)\n",
        "total_line_model = tf.keras.Model(inputs=total_lines_inputs,\n",
        "                                  outputs=y)\n",
        "\n",
        "# 5. Combine token and char embeddings into a hybrid embedding\n",
        "combined_embeddings = tf.keras.layers.Concatenate(name=\"token_char_hybrid_embedding\")([token_model.output,\n",
        "                                                                              char_model.output])\n",
        "z = tf.keras.layers.Dense(256, activation=\"relu\")(combined_embeddings)\n",
        "z = tf.keras.layers.Dropout(0.5)(z)\n",
        "\n",
        "# 6. Combine positional embeddings with combined token and char embeddings into a tribrid embedding\n",
        "z = tf.keras.layers.Concatenate(name=\"token_char_positional_embedding\")([line_number_model.output,\n",
        "                                                                total_line_model.output,\n",
        "                                                                z])\n",
        "output_layer = tf.keras.layers.Dense(5, activation=\"softmax\", name=\"output_layer\")(z)\n",
        "\n",
        "\n",
        "\n",
        "model_5 = tf.keras.Model(inputs=[line_number_model.input,\n",
        "                                 total_line_model.input,\n",
        "                                 token_model.input,\n",
        "                                 char_model.input],\n",
        "                         outputs=output_layer)"
      ],
      "metadata": {
        "id": "zkjf1ILj7mGS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_5.summary()"
      ],
      "metadata": {
        "id": "pEUe8rP_7l-5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f6d37163-22ed-41b4-ad96-e730eea32751"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_7\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                Output Shape                 Param #   Connected to                  \n",
            "==================================================================================================\n",
            " char_inputs (InputLayer)    [(None, 1)]                  0         []                            \n",
            "                                                                                                  \n",
            " token_inputs (InputLayer)   [(None,)]                    0         []                            \n",
            "                                                                                                  \n",
            " char_vectorizer (TextVecto  (None, 290)                  0         ['char_inputs[0][0]']         \n",
            " rization)                                                                                        \n",
            "                                                                                                  \n",
            " universal_sentence_encoder  (None, 512)                  2567978   ['token_inputs[0][0]']        \n",
            "  (KerasLayer)                                            24                                      \n",
            "                                                                                                  \n",
            " char_embed (Embedding)      (None, 290, 25)              1750      ['char_vectorizer[2][0]']     \n",
            "                                                                                                  \n",
            " dense_7 (Dense)             (None, 128)                  65664     ['universal_sentence_encoder[2\n",
            "                                                                    ][0]']                        \n",
            "                                                                                                  \n",
            " bidirectional_1 (Bidirecti  (None, 64)                   14848     ['char_embed[2][0]']          \n",
            " onal)                                                                                            \n",
            "                                                                                                  \n",
            " token_char_hybrid_embeddin  (None, 192)                  0         ['dense_7[0][0]',             \n",
            " g (Concatenate)                                                     'bidirectional_1[0][0]']     \n",
            "                                                                                                  \n",
            " line_number_input (InputLa  [(None, 15)]                 0         []                            \n",
            " yer)                                                                                             \n",
            "                                                                                                  \n",
            " total_lines_input (InputLa  [(None, 20)]                 0         []                            \n",
            " yer)                                                                                             \n",
            "                                                                                                  \n",
            " dense_10 (Dense)            (None, 256)                  49408     ['token_char_hybrid_embedding[\n",
            "                                                                    0][0]']                       \n",
            "                                                                                                  \n",
            " dense_8 (Dense)             (None, 32)                   512       ['line_number_input[0][0]']   \n",
            "                                                                                                  \n",
            " dense_9 (Dense)             (None, 32)                   672       ['total_lines_input[0][0]']   \n",
            "                                                                                                  \n",
            " dropout_2 (Dropout)         (None, 256)                  0         ['dense_10[0][0]']            \n",
            "                                                                                                  \n",
            " token_char_positional_embe  (None, 320)                  0         ['dense_8[0][0]',             \n",
            " dding (Concatenate)                                                 'dense_9[0][0]',             \n",
            "                                                                     'dropout_2[0][0]']           \n",
            "                                                                                                  \n",
            " output_layer (Dense)        (None, 5)                    1605      ['token_char_positional_embedd\n",
            "                                                                    ing[0][0]']                   \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 256932283 (980.12 MB)\n",
            "Trainable params: 134459 (525.23 KB)\n",
            "Non-trainable params: 256797824 (979.61 MB)\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(model_5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 856
        },
        "id": "dF7KvyWke75O",
        "outputId": "e2cf8c2d-070c-4fe8-9a24-0d834c027beb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 96
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for layer in model_5.layers:\n",
        "  print(layer, layer.trainable)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q68S8Rk9e-YQ",
        "outputId": "68d8d0b0-6dc3-4777-a6d4-a9c608ab950e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<keras.src.engine.input_layer.InputLayer object at 0x79427dc78940> True\n",
            "<keras.src.engine.input_layer.InputLayer object at 0x79427db3f460> True\n",
            "<keras.src.layers.preprocessing.text_vectorization.TextVectorization object at 0x7941e4e2f5e0> True\n",
            "<tensorflow_hub.keras_layer.KerasLayer object at 0x7942882311b0> False\n",
            "<keras.src.layers.core.embedding.Embedding object at 0x7941e4e65b70> True\n",
            "<keras.src.layers.core.dense.Dense object at 0x79427dce3e80> True\n",
            "<keras.src.layers.rnn.bidirectional.Bidirectional object at 0x79427dcbc250> True\n",
            "<keras.src.layers.merging.concatenate.Concatenate object at 0x79427db5b490> True\n",
            "<keras.src.engine.input_layer.InputLayer object at 0x79427dce0640> True\n",
            "<keras.src.engine.input_layer.InputLayer object at 0x79427db5a2c0> True\n",
            "<keras.src.layers.core.dense.Dense object at 0x79427da62440> True\n",
            "<keras.src.layers.core.dense.Dense object at 0x7941dacf91e0> True\n",
            "<keras.src.layers.core.dense.Dense object at 0x79427db075b0> True\n",
            "<keras.src.layers.regularization.dropout.Dropout object at 0x79427dbd6a10> True\n",
            "<keras.src.layers.merging.concatenate.Concatenate object at 0x79427db060b0> True\n",
            "<keras.src.layers.core.dense.Dense object at 0x79427da62ec0> True\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_5.compile(loss=tf.keras.losses.CategoricalCrossentropy(label_smoothing=0.2), # add label smoothing (examples which are really confident get smoothed a little)\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "IPKjDRD5fYK_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Create tribrid embedding datasets and fit tribrid model\n"
      ],
      "metadata": {
        "id": "p6MALjQSm0VC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create training and validation datasets (all four kinds of inputs)\n",
        "train_pos_char_token_data = tf.data.Dataset.from_tensor_slices((train_line_numbers_one_hot, # line numbers\n",
        "                                                                train_total_lines_one_hot, # total lines\n",
        "                                                                train_sentences, # train tokens\n",
        "                                                                train_chars)) # train chars\n",
        "train_pos_char_token_labels = tf.data.Dataset.from_tensor_slices(train_labels_one_hot) # train labels\n",
        "train_pos_char_token_dataset = tf.data.Dataset.zip((train_pos_char_token_data, train_pos_char_token_labels)) # combine data and labels\n",
        "train_pos_char_token_dataset = train_pos_char_token_dataset.batch(32).prefetch(tf.data.AUTOTUNE) # turn into batches and prefetch appropriately\n",
        "\n",
        "# Validation dataset\n",
        "val_pos_char_token_data = tf.data.Dataset.from_tensor_slices((val_line_numbers_one_hot,\n",
        "                                                              val_total_lines_one_hot,\n",
        "                                                              val_sentences,\n",
        "                                                              val_chars))\n",
        "val_pos_char_token_labels = tf.data.Dataset.from_tensor_slices(val_labels_one_hot)\n",
        "val_pos_char_token_dataset = tf.data.Dataset.zip((val_pos_char_token_data, val_pos_char_token_labels))\n",
        "val_pos_char_token_dataset = val_pos_char_token_dataset.batch(32).prefetch(tf.data.AUTOTUNE) # turn into batches and prefetch appropriately\n",
        "\n",
        "# Check input shapes\n",
        "train_pos_char_token_dataset, val_pos_char_token_dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "huKbKldAjyeO",
        "outputId": "6df8b375-a6f8-4ce3-c907-7b583bd1c580"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<_PrefetchDataset element_spec=((TensorSpec(shape=(None, 15), dtype=tf.float32, name=None), TensorSpec(shape=(None, 20), dtype=tf.float32, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None)), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>,\n",
              " <_PrefetchDataset element_spec=((TensorSpec(shape=(None, 15), dtype=tf.float32, name=None), TensorSpec(shape=(None, 20), dtype=tf.float32, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None)), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>)"
            ]
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "history_model_5 = model_5.fit(train_pos_char_token_dataset,\n",
        "                              steps_per_epoch=int(0.1 * len(train_pos_char_token_dataset)),\n",
        "                              epochs=3,\n",
        "                              validation_data=val_pos_char_token_dataset,\n",
        "                              validation_steps=int(0.1 * len(val_pos_char_token_dataset)))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3gpsnSfjR0a9",
        "outputId": "47aac111-c1e2-445e-c59c-ca7b71205382"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/3\n",
            "562/562 [==============================] - 28s 39ms/step - loss: 1.0961 - accuracy: 0.7233 - val_loss: 0.9804 - val_accuracy: 0.8065\n",
            "Epoch 2/3\n",
            "562/562 [==============================] - 20s 35ms/step - loss: 0.9672 - accuracy: 0.8150 - val_loss: 0.9495 - val_accuracy: 0.8238\n",
            "Epoch 3/3\n",
            "562/562 [==============================] - 20s 36ms/step - loss: 0.9513 - accuracy: 0.8227 - val_loss: 0.9389 - val_accuracy: 0.8324\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_5_pred_probs = model_5.predict(val_pos_char_token_dataset, verbose =1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iv8ctkfxR0XY",
        "outputId": "90c6ae22-fad9-4a53-97ca-3606e7344f59"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 20s 20ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_5_preds = tf.argmax(model_5_pred_probs, axis=1)\n",
        "model_5_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JJoBe1JUR0U9",
        "outputId": "748023dd-3aa8-4526-f6c6-e1b63f2101b5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(30212,), dtype=int64, numpy=array([0, 0, 3, ..., 4, 4, 1])>"
            ]
          },
          "metadata": {},
          "execution_count": 102
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_5_results = calculate_results(val_labels_encoded,\n",
        "                                    model_5_preds)\n",
        "model_5_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qaGzRQyAR0Sx",
        "outputId": "51a7f772-5d85-474e-df90-bb4fa05dd8af"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 83.15238977889581,\n",
              " 'precision': 0.8300042166207565,\n",
              " 'recall': 0.8315238977889581,\n",
              " 'f1': 0.8304523686080151}"
            ]
          },
          "metadata": {},
          "execution_count": 103
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Comparing model results"
      ],
      "metadata": {
        "id": "EtEKGPKuSqiI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "all_model_results = pd.DataFrame({\"baseline\": baseline_results,\n",
        "                                  \"custom_token_embed_conv1d\": model_1_results,\n",
        "                                  \"pretrained_token_embed\": model_2_results,\n",
        "                                  \"custom_char_embed_conv1d\": model_3_results,\n",
        "                                  \"hybrid_char_token_embed\": model_4_results,\n",
        "                                  \"tribrid_pos_char_token_embed\": model_5_results})\n",
        "all_model_results = all_model_results.transpose()\n",
        "all_model_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237
        },
        "id": "5bsbf08VR0QT",
        "outputId": "5cfe895a-1bf2-41cd-f21f-57f05a7516ca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                               accuracy  precision    recall        f1\n",
              "baseline                      72.183238   0.718647  0.721832  0.698925\n",
              "custom_token_embed_conv1d     78.809744   0.785036  0.788097  0.785680\n",
              "pretrained_token_embed        71.130677   0.711305  0.711307  0.708188\n",
              "custom_char_embed_conv1d      44.813319   0.376504  0.448133  0.394297\n",
              "hybrid_char_token_embed       73.083543   0.728908  0.730835  0.727607\n",
              "tribrid_pos_char_token_embed  83.152390   0.830004  0.831524  0.830452"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8f8ad7fb-d1cf-4f70-9ea4-15009ce95f3f\" 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>accuracy</th>\n",
              "      <th>precision</th>\n",
              "      <th>recall</th>\n",
              "      <th>f1</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>baseline</th>\n",
              "      <td>72.183238</td>\n",
              "      <td>0.718647</td>\n",
              "      <td>0.721832</td>\n",
              "      <td>0.698925</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>custom_token_embed_conv1d</th>\n",
              "      <td>78.809744</td>\n",
              "      <td>0.785036</td>\n",
              "      <td>0.788097</td>\n",
              "      <td>0.785680</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>pretrained_token_embed</th>\n",
              "      <td>71.130677</td>\n",
              "      <td>0.711305</td>\n",
              "      <td>0.711307</td>\n",
              "      <td>0.708188</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>custom_char_embed_conv1d</th>\n",
              "      <td>44.813319</td>\n",
              "      <td>0.376504</td>\n",
              "      <td>0.448133</td>\n",
              "      <td>0.394297</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>hybrid_char_token_embed</th>\n",
              "      <td>73.083543</td>\n",
              "      <td>0.728908</td>\n",
              "      <td>0.730835</td>\n",
              "      <td>0.727607</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>tribrid_pos_char_token_embed</th>\n",
              "      <td>83.152390</td>\n",
              "      <td>0.830004</td>\n",
              "      <td>0.831524</td>\n",
              "      <td>0.830452</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-8f8ad7fb-d1cf-4f70-9ea4-15009ce95f3f')\"\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-8f8ad7fb-d1cf-4f70-9ea4-15009ce95f3f 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-8f8ad7fb-d1cf-4f70-9ea4-15009ce95f3f');\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-2deda491-b3b2-418d-b1d7-c9210b98bec8\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-2deda491-b3b2-418d-b1d7-c9210b98bec8')\"\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-2deda491-b3b2-418d-b1d7-c9210b98bec8 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_975bffc3-5c1d-437a-b7d8-29543fab3e80\">\n",
              "    <style>\n",
              "      .colab-df-generate {\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-generate: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-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate: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",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('all_model_results')\"\n",
              "            title=\"Generate code using this dataframe.\"\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=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_975bffc3-5c1d-437a-b7d8-29543fab3e80 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('all_model_results');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "all_model_results[\"accuracy\"] = all_model_results[\"accuracy\"]/100"
      ],
      "metadata": {
        "id": "Sy5NQqoBR0OR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "all_model_results.plot(kind=\"bar\", figsize=(10, 7)).legend(bbox_to_anchor=(1.0, 1.0));\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 798
        },
        "id": "xgsivaWeR0L3",
        "outputId": "f0aa4c6b-bed5-4eae-fcc4-ea24b8869e6d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "all_model_results.sort_values(\"f1\", ascending=False)[\"f1\"].plot(kind=\"bar\", figsize=(10, 7))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 815
        },
        "id": "x2lm4DbIR0Jd",
        "outputId": "78b6553c-08e2-4021-aa2c-c1353a75856b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 107
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_5.save(\"skimlit_tribrid_model_10percent_20k_pubmed20k_rct\")"
      ],
      "metadata": {
        "id": "8ni_CJGWR0HH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!zip -r /content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct.zip /content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k8m9qPMxR0Es",
        "outputId": "93dd3a5a-fbdb-4668-cb85-c4390299d2c4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/ (stored 0%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/variables/ (stored 0%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/variables/variables.index (deflated 77%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/variables/variables.data-00000-of-00001 (deflated 7%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/assets/ (stored 0%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/fingerprint.pb (stored 0%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/saved_model.pb (deflated 72%)\n",
            "updating: content/skimlit_tribrid_model_10percent_20k_pubmed20k_rct/keras_metadata.pb (deflated 91%)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# model_5.save(\"skimlit_tribrid_model_10percent_20k_pubmed20k_rct.hdf5\")\n",
        "model_5.save(\"skimlit_tribrid_model_10percent_20k_pubmed20k_rct\", save_format='tf')"
      ],
      "metadata": {
        "id": "Slq1XAh_R0CN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_5.save(\"skimlit_tribrid_model_10percent_20k_pubmed20k_rct.keras\")"
      ],
      "metadata": {
        "id": "PPuYlCCiR0AP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Loading the saved model"
      ],
      "metadata": {
        "id": "zud3Yp9nedR4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "loaded_model = tf.keras.models.load_model(\"skimlit_tribrid_model_10percent_20k_pubmed20k_rct\")"
      ],
      "metadata": {
        "id": "wySSFW6MRz9r"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "loaded_model.evaluate(val_pos_char_token_dataset)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GXDcFA0ORz7P",
        "outputId": "a1816534-7731-4b40-9a40-ad9235541a54"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "945/945 [==============================] - 20s 20ms/step - loss: 0.9369 - accuracy: 0.8315\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.936908483505249, 0.8315238952636719]"
            ]
          },
          "metadata": {},
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Find most wrong"
      ],
      "metadata": {
        "id": "WlQZI_RF8Scn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_pos_char_token_data = tf.data.Dataset.from_tensor_slices((test_line_numbers_one_hot,\n",
        "                                                               test_total_lines_one_hot,\n",
        "                                                               test_sentences,\n",
        "                                                               test_chars))\n",
        "test_pos_char_token_labels = tf.data.Dataset.from_tensor_slices(test_labels_one_hot)\n",
        "test_pos_char_token_dataset = tf.data.Dataset.zip((test_pos_char_token_data, test_pos_char_token_labels))\n",
        "test_pos_char_token_dataset = test_pos_char_token_dataset.batch(32).prefetch(tf.data.AUTOTUNE)\n",
        "\n",
        "# Check shapes\n",
        "test_pos_char_token_dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "USdakEYPf2o5",
        "outputId": "13e7c2b0-4efb-4c94-ba37-e3fd756aec6b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<_PrefetchDataset element_spec=((TensorSpec(shape=(None, 15), dtype=tf.float32, name=None), TensorSpec(shape=(None, 20), dtype=tf.float32, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None), TensorSpec(shape=(None,), dtype=tf.string, name=None)), TensorSpec(shape=(None, 5), dtype=tf.float64, name=None))>"
            ]
          },
          "metadata": {},
          "execution_count": 114
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_pred_probs = loaded_model.predict(test_pos_char_token_dataset,\n",
        "                                       verbose=1)\n",
        "test_preds = tf.argmax(test_pred_probs, axis=1)\n",
        "test_preds[:10]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1WmLQETzf2mN",
        "outputId": "df13383a-99a0-4b14-ddc5-9bc294d79301"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "942/942 [==============================] - 19s 19ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=int64, numpy=array([3, 0, 2, 2, 4, 4, 4, 1, 4, 0])>"
            ]
          },
          "metadata": {},
          "execution_count": 115
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loaded_model_test_results = calculate_results(y_true=test_labels_encoded,\n",
        "                                              y_pred=test_preds)\n",
        "loaded_model_test_results\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E-1eT6rNf2jX",
        "outputId": "c2e73c46-dc3c-4dc2-bc37-dbeceffcf1e8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': 82.7775012444002,\n",
              " 'precision': 0.8261845034974992,\n",
              " 'recall': 0.827775012444002,\n",
              " 'f1': 0.8267394837462051}"
            ]
          },
          "metadata": {},
          "execution_count": 116
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Getting the most wrong"
      ],
      "metadata": {
        "id": "aK7ER2C7AEFx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%%time\n",
        "# Get list of class names of test predictions\n",
        "test_pred_classes = [label_encoder.classes_[pred] for pred in test_preds]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "87tdWF8Cf2gs",
        "outputId": "0723e656-c15b-42ad-9283-d01f65adf581"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "CPU times: user 15.1 s, sys: 788 ms, total: 15.9 s\n",
            "Wall time: 15 s\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_df[\"prediction\"] = test_pred_classes # create column with test prediction class names\n",
        "test_df[\"pred_prob\"] = tf.reduce_max(test_pred_probs, axis=1).numpy() # get the maximum prediction probability\n",
        "test_df[\"correct\"] = test_df[\"prediction\"] == test_df[\"target\"]"
      ],
      "metadata": {
        "id": "ZNbHuH01f2di"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test_df.head(20)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 676
        },
        "id": "d0vEQM72f2ax",
        "outputId": "169643fb-1ebb-4958-e9e8-aa9613b9eaea"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         target                                               text  \\\n",
              "0    BACKGROUND  this study analyzed liver function abnormaliti...   \n",
              "1       RESULTS  a post hoc analysis was conducted with the use...   \n",
              "2       RESULTS  liver function tests ( lfts ) were measured at...   \n",
              "3       RESULTS  survival analyses were used to assess the asso...   \n",
              "4       RESULTS  the percentage of patients with abnormal lfts ...   \n",
              "5       RESULTS  when mean hemodynamic profiles were compared i...   \n",
              "6       RESULTS  multivariable analyses revealed that patients ...   \n",
              "7   CONCLUSIONS  abnormal lfts are common in the adhf populatio...   \n",
              "8   CONCLUSIONS  elevated meld-xi scores are associated with po...   \n",
              "9    BACKGROUND  minimally invasive endovascular aneurysm repai...   \n",
              "10   BACKGROUND  the aim of this study was to analyse the cost-...   \n",
              "11      METHODS  resource use was determined from the amsterdam...   \n",
              "12      METHODS  the analysis was performed from a provider per...   \n",
              "13      METHODS  all costs were calculated as if all patients h...   \n",
              "14      RESULTS            a total of @ patients were randomized .   \n",
              "15      RESULTS  the @-day mortality rate was @ per cent after ...   \n",
              "16      RESULTS  at @months , the total mortality rate for evar...   \n",
              "17      RESULTS  the mean cost difference between evar and or w...   \n",
              "18      RESULTS  the incremental cost-effectiveness ratio per p...   \n",
              "19      RESULTS  there was no significant difference in quality...   \n",
              "\n",
              "    line_number  total_lines   prediction  pred_prob  correct  \n",
              "0             0            8    OBJECTIVE   0.489885    False  \n",
              "1             1            8   BACKGROUND   0.363657    False  \n",
              "2             2            8      METHODS   0.805599    False  \n",
              "3             3            8      METHODS   0.580617    False  \n",
              "4             4            8      RESULTS   0.736250     True  \n",
              "5             5            8      RESULTS   0.879160     True  \n",
              "6             6            8      RESULTS   0.557308     True  \n",
              "7             7            8  CONCLUSIONS   0.479922     True  \n",
              "8             8            8      RESULTS   0.459729    False  \n",
              "9             0           12   BACKGROUND   0.576643     True  \n",
              "10            1           12    OBJECTIVE   0.494629    False  \n",
              "11            2           12      METHODS   0.670961     True  \n",
              "12            3           12      METHODS   0.876626     True  \n",
              "13            4           12      METHODS   0.482146     True  \n",
              "14            5           12      RESULTS   0.720828     True  \n",
              "15            6           12      RESULTS   0.645031     True  \n",
              "16            7           12      RESULTS   0.896880     True  \n",
              "17            8           12      RESULTS   0.841973     True  \n",
              "18            9           12      RESULTS   0.805371     True  \n",
              "19           10           12      RESULTS   0.767485     True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a15d289d-b32b-42c7-a20d-a67eb149ccc8\" 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>target</th>\n",
              "      <th>text</th>\n",
              "      <th>line_number</th>\n",
              "      <th>total_lines</th>\n",
              "      <th>prediction</th>\n",
              "      <th>pred_prob</th>\n",
              "      <th>correct</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>this study analyzed liver function abnormaliti...</td>\n",
              "      <td>0</td>\n",
              "      <td>8</td>\n",
              "      <td>OBJECTIVE</td>\n",
              "      <td>0.489885</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>a post hoc analysis was conducted with the use...</td>\n",
              "      <td>1</td>\n",
              "      <td>8</td>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>0.363657</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>liver function tests ( lfts ) were measured at...</td>\n",
              "      <td>2</td>\n",
              "      <td>8</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.805599</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>survival analyses were used to assess the asso...</td>\n",
              "      <td>3</td>\n",
              "      <td>8</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.580617</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the percentage of patients with abnormal lfts ...</td>\n",
              "      <td>4</td>\n",
              "      <td>8</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.736250</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>when mean hemodynamic profiles were compared i...</td>\n",
              "      <td>5</td>\n",
              "      <td>8</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.879160</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>multivariable analyses revealed that patients ...</td>\n",
              "      <td>6</td>\n",
              "      <td>8</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.557308</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>abnormal lfts are common in the adhf populatio...</td>\n",
              "      <td>7</td>\n",
              "      <td>8</td>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>0.479922</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>elevated meld-xi scores are associated with po...</td>\n",
              "      <td>8</td>\n",
              "      <td>8</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.459729</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>minimally invasive endovascular aneurysm repai...</td>\n",
              "      <td>0</td>\n",
              "      <td>12</td>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>0.576643</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>the aim of this study was to analyse the cost-...</td>\n",
              "      <td>1</td>\n",
              "      <td>12</td>\n",
              "      <td>OBJECTIVE</td>\n",
              "      <td>0.494629</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>resource use was determined from the amsterdam...</td>\n",
              "      <td>2</td>\n",
              "      <td>12</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.670961</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>the analysis was performed from a provider per...</td>\n",
              "      <td>3</td>\n",
              "      <td>12</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.876626</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>all costs were calculated as if all patients h...</td>\n",
              "      <td>4</td>\n",
              "      <td>12</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.482146</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>a total of @ patients were randomized .</td>\n",
              "      <td>5</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.720828</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the @-day mortality rate was @ per cent after ...</td>\n",
              "      <td>6</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.645031</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>at @months , the total mortality rate for evar...</td>\n",
              "      <td>7</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.896880</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the mean cost difference between evar and or w...</td>\n",
              "      <td>8</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.841973</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the incremental cost-effectiveness ratio per p...</td>\n",
              "      <td>9</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.805371</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>there was no significant difference in quality...</td>\n",
              "      <td>10</td>\n",
              "      <td>12</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.767485</td>\n",
              "      <td>True</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-a15d289d-b32b-42c7-a20d-a67eb149ccc8')\"\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-a15d289d-b32b-42c7-a20d-a67eb149ccc8 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-a15d289d-b32b-42c7-a20d-a67eb149ccc8');\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-9de5fafe-e881-464a-9450-f81d50c5b1da\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-9de5fafe-e881-464a-9450-f81d50c5b1da')\"\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-9de5fafe-e881-464a-9450-f81d50c5b1da button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 119
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "top_100_wrong = test_df[test_df[\"correct\"] == False].sort_values(\"pred_prob\", ascending=False)[:100]\n",
        "top_100_wrong"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "2gr-Xkp9f2Xx",
        "outputId": "7a4532ad-e2ae-4b01-bc6a-a5f49ab2319a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            target                                               text  \\\n",
              "13874  CONCLUSIONS  symptom outcomes will be assessed and estimate...   \n",
              "2388       RESULTS  the primary endpoint is the cumulative three-y...   \n",
              "3573       RESULTS  a cluster randomised trial was implemented wit...   \n",
              "13598      METHODS  -@ % vs. fish : -@ % vs. fish + s : -@ % ; p <...   \n",
              "16347   BACKGROUND  to evaluate the effects of the lactic acid bac...   \n",
              "...            ...                                                ...   \n",
              "23592   BACKGROUND  we recorded the number of dead ducklings , gro...   \n",
              "12304   BACKGROUND  the severity of psychiatric symptoms , as meas...   \n",
              "2575   CONCLUSIONS                        clinicaltrials.gov : nct@ .   \n",
              "1220       RESULTS  the group intervention consisted of @ weekly c...   \n",
              "12456      RESULTS  icd patients were randomized @:@ to automatic ...   \n",
              "\n",
              "       line_number  total_lines  prediction  pred_prob  correct  \n",
              "13874            4            6     METHODS   0.953189    False  \n",
              "2388             4           13     METHODS   0.942299    False  \n",
              "3573             3           16     METHODS   0.934650    False  \n",
              "13598            6            9     RESULTS   0.925465    False  \n",
              "16347            0           12   OBJECTIVE   0.923894    False  \n",
              "...            ...          ...         ...        ...      ...  \n",
              "23592            4           12     METHODS   0.843512    False  \n",
              "12304            2           16     METHODS   0.843202    False  \n",
              "2575            10           10  BACKGROUND   0.843010    False  \n",
              "1220             2           13     METHODS   0.842704    False  \n",
              "12456            1           13     METHODS   0.842024    False  \n",
              "\n",
              "[100 rows x 7 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-83696520-d05c-49dc-9990-15e6ff104179\" 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>target</th>\n",
              "      <th>text</th>\n",
              "      <th>line_number</th>\n",
              "      <th>total_lines</th>\n",
              "      <th>prediction</th>\n",
              "      <th>pred_prob</th>\n",
              "      <th>correct</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13874</th>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>symptom outcomes will be assessed and estimate...</td>\n",
              "      <td>4</td>\n",
              "      <td>6</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.953189</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2388</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the primary endpoint is the cumulative three-y...</td>\n",
              "      <td>4</td>\n",
              "      <td>13</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.942299</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3573</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>a cluster randomised trial was implemented wit...</td>\n",
              "      <td>3</td>\n",
              "      <td>16</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.934650</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13598</th>\n",
              "      <td>METHODS</td>\n",
              "      <td>-@ % vs. fish : -@ % vs. fish + s : -@ % ; p &lt;...</td>\n",
              "      <td>6</td>\n",
              "      <td>9</td>\n",
              "      <td>RESULTS</td>\n",
              "      <td>0.925465</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16347</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>to evaluate the effects of the lactic acid bac...</td>\n",
              "      <td>0</td>\n",
              "      <td>12</td>\n",
              "      <td>OBJECTIVE</td>\n",
              "      <td>0.923894</td>\n",
              "      <td>False</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23592</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>we recorded the number of dead ducklings , gro...</td>\n",
              "      <td>4</td>\n",
              "      <td>12</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.843512</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12304</th>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>the severity of psychiatric symptoms , as meas...</td>\n",
              "      <td>2</td>\n",
              "      <td>16</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.843202</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2575</th>\n",
              "      <td>CONCLUSIONS</td>\n",
              "      <td>clinicaltrials.gov : nct@ .</td>\n",
              "      <td>10</td>\n",
              "      <td>10</td>\n",
              "      <td>BACKGROUND</td>\n",
              "      <td>0.843010</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1220</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>the group intervention consisted of @ weekly c...</td>\n",
              "      <td>2</td>\n",
              "      <td>13</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.842704</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12456</th>\n",
              "      <td>RESULTS</td>\n",
              "      <td>icd patients were randomized @:@ to automatic ...</td>\n",
              "      <td>1</td>\n",
              "      <td>13</td>\n",
              "      <td>METHODS</td>\n",
              "      <td>0.842024</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>100 rows × 7 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-83696520-d05c-49dc-9990-15e6ff104179')\"\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-83696520-d05c-49dc-9990-15e6ff104179 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-83696520-d05c-49dc-9990-15e6ff104179');\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-5096eac1-7b81-4720-8da0-e01cb9eccd1f\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5096eac1-7b81-4720-8da0-e01cb9eccd1f')\"\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-5096eac1-7b81-4720-8da0-e01cb9eccd1f button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_5455ea9a-bd2b-4da6-b2a6-c1dd5a92ab0a\">\n",
              "    <style>\n",
              "      .colab-df-generate {\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-generate: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-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate: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",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('top_100_wrong')\"\n",
              "            title=\"Generate code using this dataframe.\"\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=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_5455ea9a-bd2b-4da6-b2a6-c1dd5a92ab0a button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('top_100_wrong');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 120
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for row in top_100_wrong[0:10].itertuples(): # adjust indexes to view different samples\n",
        "  _, target, text, line_number, total_lines, prediction, pred_prob, _ = row\n",
        "  print(f\"Target: {target}, Pred: {prediction}, Prob: {pred_prob}, Line number: {line_number}, Total lines: {total_lines}\\n\")\n",
        "  print(f\"Text:\\n{text}\\n\")\n",
        "  print(\"-----\\n\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YGuTF_wujk4_",
        "outputId": "59c5e2f7-7a9c-4b2b-9327-e6fe7005f337"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Target: CONCLUSIONS, Pred: METHODS, Prob: 0.9531888365745544, Line number: 4, Total lines: 6\n",
            "\n",
            "Text:\n",
            "symptom outcomes will be assessed and estimates of cost-effectiveness made .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: RESULTS, Pred: METHODS, Prob: 0.9422993659973145, Line number: 4, Total lines: 13\n",
            "\n",
            "Text:\n",
            "the primary endpoint is the cumulative three-year hiv incidence .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: RESULTS, Pred: METHODS, Prob: 0.9346501231193542, Line number: 3, Total lines: 16\n",
            "\n",
            "Text:\n",
            "a cluster randomised trial was implemented with @,@ children in @ government primary schools on the south coast of kenya in @-@ .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: METHODS, Pred: RESULTS, Prob: 0.9254651665687561, Line number: 6, Total lines: 9\n",
            "\n",
            "Text:\n",
            "-@ % vs. fish : -@ % vs. fish + s : -@ % ; p < @ ) but there were no significant differences between groups .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: BACKGROUND, Pred: OBJECTIVE, Prob: 0.9238936901092529, Line number: 0, Total lines: 12\n",
            "\n",
            "Text:\n",
            "to evaluate the effects of the lactic acid bacterium lactobacillus salivarius on caries risk factors .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: METHODS, Pred: BACKGROUND, Prob: 0.9200336337089539, Line number: 1, Total lines: 11\n",
            "\n",
            "Text:\n",
            "pretest-posttest .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: BACKGROUND, Pred: OBJECTIVE, Prob: 0.9038149118423462, Line number: 0, Total lines: 10\n",
            "\n",
            "Text:\n",
            "to validate the association between accommodation and visual asthenopia by measuring objective accommodative amplitude with the optical quality analysis system ( oqas , visiometrics , terrassa , spain ) , and to investigate associations among accommodation , ocular surface instability , and visual asthenopia while viewing @d displays .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: RESULTS, Pred: METHODS, Prob: 0.9029855728149414, Line number: 4, Total lines: 14\n",
            "\n",
            "Text:\n",
            "a screening questionnaire for moh was sent to all @-@ year old patients on these gps ` list .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: RESULTS, Pred: METHODS, Prob: 0.9019884467124939, Line number: 6, Total lines: 15\n",
            "\n",
            "Text:\n",
            "analysis was by linear mixed model .\n",
            "\n",
            "-----\n",
            "\n",
            "Target: RESULTS, Pred: METHODS, Prob: 0.901593804359436, Line number: 3, Total lines: 6\n",
            "\n",
            "Text:\n",
            "repeated measures analysis of variance ( anova ) tests were conducted to examine the differences across conditions and time .\n",
            "\n",
            "-----\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Make example predictions\n",
        "\n"
      ],
      "metadata": {
        "id": "ojxpxdzQCbTk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import json\n",
        "\n",
        "!wget https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/main/extras/skimlit_example_abstracts.json\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "movP8Yhrjk1Y",
        "outputId": "979b28fc-d8bb-4d88-af5d-1237157e81bc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2024-01-04 06:49:33--  https://raw.githubusercontent.com/mrdbourke/tensorflow-deep-learning/main/extras/skimlit_example_abstracts.json\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 6737 (6.6K) [text/plain]\n",
            "Saving to: ‘skimlit_example_abstracts.json.1’\n",
            "\n",
            "skimlit_example_abs 100%[===================>]   6.58K  --.-KB/s    in 0s      \n",
            "\n",
            "2024-01-04 06:49:34 (92.6 MB/s) - ‘skimlit_example_abstracts.json.1’ saved [6737/6737]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "with open(\"skimlit_example_abstracts.json\", \"r\") as f:\n",
        "  example_abstracts = json.load(f)\n",
        "\n",
        "example_abstracts\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ay-13637jkyj",
        "outputId": "feb981eb-31b3-4d37-8dc9-aaa782a7607f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[{'abstract': 'This RCT examined the efficacy of a manualized social intervention for children with HFASDs. Participants were randomly assigned to treatment or wait-list conditions. Treatment included instruction and therapeutic activities targeting social skills, face-emotion recognition, interest expansion, and interpretation of non-literal language. A response-cost program was applied to reduce problem behaviors and foster skills acquisition. Significant treatment effects were found for five of seven primary outcome measures (parent ratings and direct child measures). Secondary measures based on staff ratings (treatment group only) corroborated gains reported by parents. High levels of parent, child and staff satisfaction were reported, along with high levels of treatment fidelity. Standardized effect size estimates were primarily in the medium and large ranges and favored the treatment group.',\n",
              "  'source': 'https://pubmed.ncbi.nlm.nih.gov/20232240/',\n",
              "  'details': 'RCT of a manualized social treatment for high-functioning autism spectrum disorders'},\n",
              " {'abstract': \"Postpartum depression (PPD) is the most prevalent mood disorder associated with childbirth. No single cause of PPD has been identified, however the increased risk of nutritional deficiencies incurred through the high nutritional requirements of pregnancy may play a role in the pathology of depressive symptoms. Three nutritional interventions have drawn particular interest as possible non-invasive and cost-effective prevention and/or treatment strategies for PPD; omega-3 (n-3) long chain polyunsaturated fatty acids (LCPUFA), vitamin D and overall diet. We searched for meta-analyses of randomised controlled trials (RCT's) of nutritional interventions during the perinatal period with PPD as an outcome, and checked for any trials published subsequently to the meta-analyses. Fish oil: Eleven RCT's of prenatal fish oil supplementation RCT's show null and positive effects on PPD symptoms. Vitamin D: no relevant RCT's were identified, however seven observational studies of maternal vitamin D levels with PPD outcomes showed inconsistent associations. Diet: Two Australian RCT's with dietary advice interventions in pregnancy had a positive and null result on PPD. With the exception of fish oil, few RCT's with nutritional interventions during pregnancy assess PPD. Further research is needed to determine whether nutritional intervention strategies during pregnancy can protect against symptoms of PPD. Given the prevalence of PPD and ease of administering PPD measures, we recommend future prenatal nutritional RCT's include PPD as an outcome.\",\n",
              "  'source': 'https://pubmed.ncbi.nlm.nih.gov/28012571/',\n",
              "  'details': 'Formatting removed (can be used to compare model to actual example)'},\n",
              " {'abstract': 'Mental illness, including depression, anxiety and bipolar disorder, accounts for a significant proportion of global disability and poses a substantial social, economic and heath burden. Treatment is presently dominated by pharmacotherapy, such as antidepressants, and psychotherapy, such as cognitive behavioural therapy; however, such treatments avert less than half of the disease burden, suggesting that additional strategies are needed to prevent and treat mental disorders. There are now consistent mechanistic, observational and interventional data to suggest diet quality may be a modifiable risk factor for mental illness. This review provides an overview of the nutritional psychiatry field. It includes a discussion of the neurobiological mechanisms likely modulated by diet, the use of dietary and nutraceutical interventions in mental disorders, and recommendations for further research. Potential biological pathways related to mental disorders include inflammation, oxidative stress, the gut microbiome, epigenetic modifications and neuroplasticity. Consistent epidemiological evidence, particularly for depression, suggests an association between measures of diet quality and mental health, across multiple populations and age groups; these do not appear to be explained by other demographic, lifestyle factors or reverse causality. Our recently published intervention trial provides preliminary clinical evidence that dietary interventions in clinically diagnosed populations are feasible and can provide significant clinical benefit. Furthermore, nutraceuticals including n-3 fatty acids, folate, S-adenosylmethionine, N-acetyl cysteine and probiotics, among others, are promising avenues for future research. Continued research is now required to investigate the efficacy of intervention studies in large cohorts and within clinically relevant populations, particularly in patients with schizophrenia, bipolar and anxiety disorders.',\n",
              "  'source': 'https://pubmed.ncbi.nlm.nih.gov/28942748/',\n",
              "  'details': 'Effect of nutrition on mental health'},\n",
              " {'abstract': \"Hepatitis C virus (HCV) and alcoholic liver disease (ALD), either alone or in combination, count for more than two thirds of all liver diseases in the Western world. There is no safe level of drinking in HCV-infected patients and the most effective goal for these patients is total abstinence. Baclofen, a GABA(B) receptor agonist, represents a promising pharmacotherapy for alcohol dependence (AD). Previously, we performed a randomized clinical trial (RCT), which demonstrated the safety and efficacy of baclofen in patients affected by AD and cirrhosis. The goal of this post-hoc analysis was to explore baclofen's effect in a subgroup of alcohol-dependent HCV-infected cirrhotic patients. Any patient with HCV infection was selected for this analysis. Among the 84 subjects randomized in the main trial, 24 alcohol-dependent cirrhotic patients had a HCV infection; 12 received baclofen 10mg t.i.d. and 12 received placebo for 12-weeks. With respect to the placebo group (3/12, 25.0%), a significantly higher number of patients who achieved and maintained total alcohol abstinence was found in the baclofen group (10/12, 83.3%; p=0.0123). Furthermore, in the baclofen group, compared to placebo, there was a significantly higher increase in albumin values from baseline (p=0.0132) and a trend toward a significant reduction in INR levels from baseline (p=0.0716). In conclusion, baclofen was safe and significantly more effective than placebo in promoting alcohol abstinence, and improving some Liver Function Tests (LFTs) (i.e. albumin, INR) in alcohol-dependent HCV-infected cirrhotic patients. Baclofen may represent a clinically relevant alcohol pharmacotherapy for these patients.\",\n",
              "  'source': 'https://pubmed.ncbi.nlm.nih.gov/22244707/',\n",
              "  'details': 'Baclofen promotes alcohol abstinence in alcohol dependent cirrhotic patients with hepatitis C virus (HCV) infection'}]"
            ]
          },
          "metadata": {},
          "execution_count": 123
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "abstracts = pd.DataFrame(example_abstracts)\n",
        "abstracts"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 174
        },
        "id": "wusvPeG5CW3g",
        "outputId": "908ca51a-2586-4fe8-c003-e2ba2a679a56"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                            abstract  \\\n",
              "0  This RCT examined the efficacy of a manualized...   \n",
              "1  Postpartum depression (PPD) is the most preval...   \n",
              "2  Mental illness, including depression, anxiety ...   \n",
              "3  Hepatitis C virus (HCV) and alcoholic liver di...   \n",
              "\n",
              "                                      source  \\\n",
              "0  https://pubmed.ncbi.nlm.nih.gov/20232240/   \n",
              "1  https://pubmed.ncbi.nlm.nih.gov/28012571/   \n",
              "2  https://pubmed.ncbi.nlm.nih.gov/28942748/   \n",
              "3  https://pubmed.ncbi.nlm.nih.gov/22244707/   \n",
              "\n",
              "                                             details  \n",
              "0  RCT of a manualized social treatment for high-...  \n",
              "1  Formatting removed (can be used to compare mod...  \n",
              "2               Effect of nutrition on mental health  \n",
              "3  Baclofen promotes alcohol abstinence in alcoho...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b63e12c4-d254-463a-a69f-9a4649b10347\" 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>abstract</th>\n",
              "      <th>source</th>\n",
              "      <th>details</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>This RCT examined the efficacy of a manualized...</td>\n",
              "      <td>https://pubmed.ncbi.nlm.nih.gov/20232240/</td>\n",
              "      <td>RCT of a manualized social treatment for high-...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Postpartum depression (PPD) is the most preval...</td>\n",
              "      <td>https://pubmed.ncbi.nlm.nih.gov/28012571/</td>\n",
              "      <td>Formatting removed (can be used to compare mod...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Mental illness, including depression, anxiety ...</td>\n",
              "      <td>https://pubmed.ncbi.nlm.nih.gov/28942748/</td>\n",
              "      <td>Effect of nutrition on mental health</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Hepatitis C virus (HCV) and alcoholic liver di...</td>\n",
              "      <td>https://pubmed.ncbi.nlm.nih.gov/22244707/</td>\n",
              "      <td>Baclofen promotes alcohol abstinence in alcoho...</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-b63e12c4-d254-463a-a69f-9a4649b10347')\"\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-b63e12c4-d254-463a-a69f-9a4649b10347 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-b63e12c4-d254-463a-a69f-9a4649b10347');\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-dcd7b521-ef06-43d2-8523-bf839bd437c1\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-dcd7b521-ef06-43d2-8523-bf839bd437c1')\"\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-dcd7b521-ef06-43d2-8523-bf839bd437c1 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_92434c8e-32ab-4bc2-9ee4-d4fb940adc4f\">\n",
              "    <style>\n",
              "      .colab-df-generate {\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-generate: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-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate: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",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('abstracts')\"\n",
              "            title=\"Generate code using this dataframe.\"\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=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_92434c8e-32ab-4bc2-9ee4-d4fb940adc4f button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('abstracts');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 124
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create sentencizer - Source: https://spacy.io/usage/linguistic-features#sbd\n",
        "from spacy.lang.en import English\n",
        "nlp = English() # setup English sentence parser\n",
        "sentencizer = nlp.add_pipe(\"sentencizer\") # create sentence splitting pipeline object\n",
        "\n",
        "doc = nlp(example_abstracts[0][\"abstract\"])\n",
        "abstract_lines = [str(sent) for sent in list(doc.sents)] # return detected sentences from doc in string type (not spaCy token type)\n",
        "abstract_lines"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zL3sfIdtCW0T",
        "outputId": "109798c5-fe69-4738-a71a-b5b1d57d4643"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['This RCT examined the efficacy of a manualized social intervention for children with HFASDs.',\n",
              " 'Participants were randomly assigned to treatment or wait-list conditions.',\n",
              " 'Treatment included instruction and therapeutic activities targeting social skills, face-emotion recognition, interest expansion, and interpretation of non-literal language.',\n",
              " 'A response-cost program was applied to reduce problem behaviors and foster skills acquisition.',\n",
              " 'Significant treatment effects were found for five of seven primary outcome measures (parent ratings and direct child measures).',\n",
              " 'Secondary measures based on staff ratings (treatment group only) corroborated gains reported by parents.',\n",
              " 'High levels of parent, child and staff satisfaction were reported, along with high levels of treatment fidelity.',\n",
              " 'Standardized effect size estimates were primarily in the medium and large ranges and favored the treatment group.']"
            ]
          },
          "metadata": {},
          "execution_count": 125
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "total_lines_in_sample = len(abstract_lines)\n",
        "\n",
        "sample_lines = []\n",
        "for i, line in enumerate(abstract_lines):\n",
        "  sample_dict = {}\n",
        "  sample_dict[\"text\"] = str(line)\n",
        "  sample_dict[\"line_number\"] = i\n",
        "  sample_dict[\"total_lines\"] = total_lines_in_sample - 1\n",
        "  sample_lines.append(sample_dict)\n",
        "sample_lines"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b_PuquJXCWw7",
        "outputId": "3e117bc0-f883-4e19-9e09-8b05540f344c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[{'text': 'This RCT examined the efficacy of a manualized social intervention for children with HFASDs.',\n",
              "  'line_number': 0,\n",
              "  'total_lines': 7},\n",
              " {'text': 'Participants were randomly assigned to treatment or wait-list conditions.',\n",
              "  'line_number': 1,\n",
              "  'total_lines': 7},\n",
              " {'text': 'Treatment included instruction and therapeutic activities targeting social skills, face-emotion recognition, interest expansion, and interpretation of non-literal language.',\n",
              "  'line_number': 2,\n",
              "  'total_lines': 7},\n",
              " {'text': 'A response-cost program was applied to reduce problem behaviors and foster skills acquisition.',\n",
              "  'line_number': 3,\n",
              "  'total_lines': 7},\n",
              " {'text': 'Significant treatment effects were found for five of seven primary outcome measures (parent ratings and direct child measures).',\n",
              "  'line_number': 4,\n",
              "  'total_lines': 7},\n",
              " {'text': 'Secondary measures based on staff ratings (treatment group only) corroborated gains reported by parents.',\n",
              "  'line_number': 5,\n",
              "  'total_lines': 7},\n",
              " {'text': 'High levels of parent, child and staff satisfaction were reported, along with high levels of treatment fidelity.',\n",
              "  'line_number': 6,\n",
              "  'total_lines': 7},\n",
              " {'text': 'Standardized effect size estimates were primarily in the medium and large ranges and favored the treatment group.',\n",
              "  'line_number': 7,\n",
              "  'total_lines': 7}]"
            ]
          },
          "metadata": {},
          "execution_count": 126
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_abstract_line_numbers = [line[\"line_number\"] for line in sample_lines]\n",
        "# One-hot encode to same depth as training data, so model accepts right input shape\n",
        "test_abstract_line_numbers_one_hot = tf.one_hot(test_abstract_line_numbers, depth=15)\n",
        "test_abstract_line_numbers_one_hot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rOk9FOf3CWt1",
        "outputId": "42ae1225-edf2-45d8-f247-f47c72b9479a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(8, 15), dtype=float32, numpy=\n",
              "array([[1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.]],\n",
              "      dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 127
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_abstract_total_lines = [line[\"total_lines\"] for line in sample_lines]\n",
        "# One-hot encode to same depth as training data, so model accepts right input shape\n",
        "test_abstract_total_lines_one_hot = tf.one_hot(test_abstract_total_lines, depth=20)\n",
        "test_abstract_total_lines_one_hot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2tgjxkeTCWqU",
        "outputId": "522007b1-625c-4441-93e2-95ea4feafd4e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(8, 20), dtype=float32, numpy=\n",
              "array([[0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.],\n",
              "       [0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
              "        0., 0., 0., 0.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 128
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "abstract_chars = [split_chars(sentence) for sentence in abstract_lines]\n",
        "abstract_chars"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-yxbAvTlCWmP",
        "outputId": "46eb7e9f-09c4-4f0d-e7f5-b310e9b3b7b1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['T h i s   R C T   e x a m i n e d   t h e   e f f i c a c y   o f   a   m a n u a l i z e d   s o c i a l   i n t e r v e n t i o n   f o r   c h i l d r e n   w i t h   H F A S D s .',\n",
              " 'P a r t i c i p a n t s   w e r e   r a n d o m l y   a s s i g n e d   t o   t r e a t m e n t   o r   w a i t - l i s t   c o n d i t i o n s .',\n",
              " 'T r e a t m e n t   i n c l u d e d   i n s t r u c t i o n   a n d   t h e r a p e u t i c   a c t i v i t i e s   t a r g e t i n g   s o c i a l   s k i l l s ,   f a c e - e m o t i o n   r e c o g n i t i o n ,   i n t e r e s t   e x p a n s i o n ,   a n d   i n t e r p r e t a t i o n   o f   n o n - l i t e r a l   l a n g u a g e .',\n",
              " 'A   r e s p o n s e - c o s t   p r o g r a m   w a s   a p p l i e d   t o   r e d u c e   p r o b l e m   b e h a v i o r s   a n d   f o s t e r   s k i l l s   a c q u i s i t i o n .',\n",
              " 'S i g n i f i c a n t   t r e a t m e n t   e f f e c t s   w e r e   f o u n d   f o r   f i v e   o f   s e v e n   p r i m a r y   o u t c o m e   m e a s u r e s   ( p a r e n t   r a t i n g s   a n d   d i r e c t   c h i l d   m e a s u r e s ) .',\n",
              " 'S e c o n d a r y   m e a s u r e s   b a s e d   o n   s t a f f   r a t i n g s   ( t r e a t m e n t   g r o u p   o n l y )   c o r r o b o r a t e d   g a i n s   r e p o r t e d   b y   p a r e n t s .',\n",
              " 'H i g h   l e v e l s   o f   p a r e n t ,   c h i l d   a n d   s t a f f   s a t i s f a c t i o n   w e r e   r e p o r t e d ,   a l o n g   w i t h   h i g h   l e v e l s   o f   t r e a t m e n t   f i d e l i t y .',\n",
              " 'S t a n d a r d i z e d   e f f e c t   s i z e   e s t i m a t e s   w e r e   p r i m a r i l y   i n   t h e   m e d i u m   a n d   l a r g e   r a n g e s   a n d   f a v o r e d   t h e   t r e a t m e n t   g r o u p .']"
            ]
          },
          "metadata": {},
          "execution_count": 129
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# Make predictions on sample abstract features\n",
        "%%time\n",
        "test_abstract_pred_probs = loaded_model.predict(x=(test_abstract_line_numbers_one_hot,\n",
        "                                                   test_abstract_total_lines_one_hot,\n",
        "                                                   tf.constant(abstract_lines),\n",
        "                                                   tf.constant(abstract_chars)))\n",
        "test_abstract_pred_probs"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5d-hiaTGCWi8",
        "outputId": "df6ebcc3-92b5-4209-8a5f-b079bfb7c095"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 2s 2s/step\n",
            "CPU times: user 1.66 s, sys: 15.8 ms, total: 1.67 s\n",
            "Wall time: 1.71 s\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.26823732, 0.09101655, 0.01682123, 0.5934786 , 0.03044627],\n",
              "       [0.06494986, 0.02460679, 0.76638836, 0.08508575, 0.05896924],\n",
              "       [0.12331747, 0.0566563 , 0.54746395, 0.18812805, 0.08443418],\n",
              "       [0.10273596, 0.13145894, 0.5140466 , 0.08428603, 0.16747247],\n",
              "       [0.05352934, 0.10930076, 0.48644873, 0.0465278 , 0.30419332],\n",
              "       [0.02997833, 0.11883163, 0.55795974, 0.04052654, 0.25270373],\n",
              "       [0.02849626, 0.12828703, 0.06387581, 0.02398072, 0.7553602 ],\n",
              "       [0.01986041, 0.13917887, 0.27850503, 0.03192266, 0.530533  ]],\n",
              "      dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 130
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_abstract_preds = tf.argmax(test_abstract_pred_probs, axis=1)\n",
        "test_abstract_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GV9svgnGCWff",
        "outputId": "87346dc2-8305-4b5c-de41-972213c68410"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(8,), dtype=int64, numpy=array([3, 2, 2, 2, 2, 2, 4, 4])>"
            ]
          },
          "metadata": {},
          "execution_count": 131
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_abstract_pred_classes = [label_encoder.classes_[i] for i in test_abstract_preds]\n",
        "test_abstract_pred_classes"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xs1XQqGZCE8O",
        "outputId": "773c84aa-ea65-41ef-c405-107decfd89e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['OBJECTIVE',\n",
              " 'METHODS',\n",
              " 'METHODS',\n",
              " 'METHODS',\n",
              " 'METHODS',\n",
              " 'METHODS',\n",
              " 'RESULTS',\n",
              " 'RESULTS']"
            ]
          },
          "metadata": {},
          "execution_count": 132
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for i, line in enumerate(abstract_lines):\n",
        "  print(f\"{test_abstract_pred_classes[i]}: {line}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6cjnLinkCLQX",
        "outputId": "2bc097b6-4934-4ffc-a85b-83ca726722c2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "OBJECTIVE: This RCT examined the efficacy of a manualized social intervention for children with HFASDs.\n",
            "METHODS: Participants were randomly assigned to treatment or wait-list conditions.\n",
            "METHODS: Treatment included instruction and therapeutic activities targeting social skills, face-emotion recognition, interest expansion, and interpretation of non-literal language.\n",
            "METHODS: A response-cost program was applied to reduce problem behaviors and foster skills acquisition.\n",
            "METHODS: Significant treatment effects were found for five of seven primary outcome measures (parent ratings and direct child measures).\n",
            "METHODS: Secondary measures based on staff ratings (treatment group only) corroborated gains reported by parents.\n",
            "RESULTS: High levels of parent, child and staff satisfaction were reported, along with high levels of treatment fidelity.\n",
            "RESULTS: Standardized effect size estimates were primarily in the medium and large ranges and favored the treatment group.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Model - 6 (Training on the 100% 20k pubmedrct data)"
      ],
      "metadata": {
        "id": "7RQZyZjmEbJc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 1. Token inputs\n",
        "token_inputs = tf.keras.layers.Input(shape=[], dtype=tf.string, name=\"token_inputs\")\n",
        "token_embeddings = tf_hub_embedding_layer(token_inputs)\n",
        "token_outputs = tf.keras.layers.Dense(128, activation=\"relu\")(token_embeddings)\n",
        "token_model = tf.keras.Model(token_inputs, token_outputs)\n",
        "\n",
        "# 2. Character inputs\n",
        "char_inputs = tf.keras.layers.Input(shape=(1,), dtype=tf.string, name=\"char_inputs\")\n",
        "char_vectors = char_vectorizer(char_inputs)\n",
        "char_embeddings = char_embed(char_vectors)\n",
        "char_bi_lstm = tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32))(char_embeddings)\n",
        "char_model = tf.keras.Model(char_inputs, char_bi_lstm)\n",
        "\n",
        "# 3. Line numbers inputs\n",
        "line_number_inputs = tf.keras.layers.Input(shape=(15,), dtype=tf.int32, name=\"line_number_input\")\n",
        "x = tf.keras.layers.Dense(32, activation=\"relu\")(line_number_inputs)\n",
        "line_number_model = tf.keras.Model(inputs=line_number_inputs,\n",
        "                                   outputs=x)\n",
        "\n",
        "# 4. Total lines inputs\n",
        "total_lines_inputs = tf.keras.layers.Input(shape=(20,), dtype=tf.int32, name=\"total_lines_input\")\n",
        "y = tf.keras.layers.Dense(32, activation=\"relu\")(total_lines_inputs)\n",
        "total_line_model = tf.keras.Model(inputs=total_lines_inputs,\n",
        "                                  outputs=y)\n",
        "\n",
        "# 5. Combine token and char embeddings into a hybrid embedding\n",
        "combined_embeddings = tf.keras.layers.Concatenate(name=\"token_char_hybrid_embedding\")([token_model.output,\n",
        "                                                                              char_model.output])\n",
        "z = tf.keras.layers.Dense(256, activation=\"relu\")(combined_embeddings)\n",
        "z = tf.keras.layers.Dropout(0.5)(z)\n",
        "\n",
        "# 6. Combine positional embeddings with combined token and char embeddings into a tribrid embedding\n",
        "z = tf.keras.layers.Concatenate(name=\"token_char_positional_embedding\")([line_number_model.output,\n",
        "                                                                total_line_model.output,\n",
        "                                                                z])\n",
        "output_layer = tf.keras.layers.Dense(5, activation=\"softmax\", name=\"output_layer\")(z)\n",
        "\n",
        "\n",
        "\n",
        "model_6 = tf.keras.Model(inputs=[line_number_model.input,\n",
        "                                 total_line_model.input,\n",
        "                                 token_model.input,\n",
        "                                 char_model.input],\n",
        "                         outputs=output_layer)"
      ],
      "metadata": {
        "id": "9iAW2_qzRDeF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(model_6)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 856
        },
        "id": "3qCo2GoDRbwi",
        "outputId": "4895501f-65a9-42d0-a60a-9fb81f288863"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 146
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_6.compile(loss=tf.keras.losses.BinaryCrossentropy(label_smoothing=0.2),\n",
        "                optimizer=tf.keras.optimizers.Adam(),\n",
        "                metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "i3zzxxriEtlP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Creating model callbacks\n",
        "early_stopping = tf.keras.callbacks.EarlyStopping(monitor=\"val_loss\",\n",
        "                                                  patience = 3)\n",
        "checkpoint_path = \"ckpt/\"\n",
        "model_checkpoint = tf.keras.callbacks.ModelCheckpoint(checkpoint_path,\n",
        "                                                      save_best_only = True,\n",
        "                                                      monitor=\"val_loss\")"
      ],
      "metadata": {
        "id": "DmrD1bkeIW6-"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history_model_6 = model_6.fit(train_pos_char_token_dataset,\n",
        "                              steps_per_epoch=int(len(train_pos_char_token_dataset)),\n",
        "                              epochs=100,\n",
        "                              validation_data=val_pos_char_token_dataset,\n",
        "                              validation_steps=int(len(val_pos_char_token_dataset)),\n",
        "                              callbacks=[early_stopping, model_checkpoint])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P2Pad1lhH4yp",
        "outputId": "c44c3401-b320-4335-d598-15ef5e38114f"
      },
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "5627/5627 [==============================] - 223s 40ms/step - loss: 0.4176 - accuracy: 0.8174 - val_loss: 0.4051 - val_accuracy: 0.8397\n",
            "Epoch 2/100\n",
            "5627/5627 [==============================] - 229s 41ms/step - loss: 0.4058 - accuracy: 0.8439 - val_loss: 0.4029 - val_accuracy: 0.8433\n",
            "Epoch 3/100\n",
            "5627/5627 [==============================] - 221s 39ms/step - loss: 0.4019 - accuracy: 0.8536 - val_loss: 0.4016 - val_accuracy: 0.8462\n",
            "Epoch 4/100\n",
            "5627/5627 [==============================] - 236s 42ms/step - loss: 0.3993 - accuracy: 0.8604 - val_loss: 0.4012 - val_accuracy: 0.8454\n",
            "Epoch 5/100\n",
            "5627/5627 [==============================] - 199s 35ms/step - loss: 0.3971 - accuracy: 0.8659 - val_loss: 0.4014 - val_accuracy: 0.8447\n",
            "Epoch 6/100\n",
            "5627/5627 [==============================] - 199s 35ms/step - loss: 0.3954 - accuracy: 0.8698 - val_loss: 0.4021 - val_accuracy: 0.8442\n",
            "Epoch 7/100\n",
            "5627/5627 [==============================] - 191s 34ms/step - loss: 0.3939 - accuracy: 0.8741 - val_loss: 0.4021 - val_accuracy: 0.8430\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "FD44ozV7Ly1W"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}