Emotion Detection using fine-tuned BERT.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing all the necessary libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2022-12-13T18:00:58.491207Z",
     "iopub.status.busy": "2022-12-13T18:00:58.490823Z",
     "iopub.status.idle": "2022-12-13T18:00:58.501226Z",
     "shell.execute_reply": "2022-12-13T18:00:58.500003Z",
     "shell.execute_reply.started": "2022-12-13T18:00:58.491156Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/emotions-dataset-for-nlp/val.txt\n",
      "/kaggle/input/emotions-dataset-for-nlp/test.txt\n",
      "/kaggle/input/emotions-dataset-for-nlp/train.txt\n"
     ]
    }
   ],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # df processing, CSV file I/O (e.g. pd.read_csv)\n",
    "\n",
    "# Input df files are available in the read-only \"../input/\" directory\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "import os\n",
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))\n",
    "\n",
    "# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n",
    "# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:00:58.504518Z",
     "iopub.status.busy": "2022-12-13T18:00:58.503535Z",
     "iopub.status.idle": "2022-12-13T18:00:59.829690Z",
     "shell.execute_reply": "2022-12-13T18:00:59.828380Z",
     "shell.execute_reply.started": "2022-12-13T18:00:58.504482Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
      "To disable this warning, you can either:\n",
      "\t- Avoid using `tokenizers` before the fork if possible\n",
      "\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
      "--2022-12-13 18:00:59--  https://raw.githubusercontent.com/yogawicaksana/helper_prabowo/main/helper_prabowo_ml.py\n",
      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.110.133, 185.199.109.133, 185.199.108.133, ...\n",
      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.110.133|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 13881 (14K) [text/plain]\n",
      "Saving to: ‘helper_prabowo_ml.py.9’\n",
      "\n",
      "helper_prabowo_ml.p 100%[===================>]  13.56K  --.-KB/s    in 0.001s  \n",
      "\n",
      "2022-12-13 18:00:59 (20.4 MB/s) - ‘helper_prabowo_ml.py.9’ saved [13881/13881]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget https://raw.githubusercontent.com/yogawicaksana/helper_prabowo/main/helper_prabowo_ml.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:00:59.837822Z",
     "iopub.status.busy": "2022-12-13T18:00:59.832809Z",
     "iopub.status.idle": "2022-12-13T18:01:10.910876Z",
     "shell.execute_reply": "2022-12-13T18:01:10.909610Z",
     "shell.execute_reply.started": "2022-12-13T18:00:59.837778Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
      "To disable this warning, you can either:\n",
      "\t- Avoid using `tokenizers` before the fork if possible\n",
      "\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
      "Requirement already satisfied: transformers in /opt/conda/lib/python3.7/site-packages (4.20.1)\n",
      "Requirement already satisfied: pyyaml>=5.1 in /opt/conda/lib/python3.7/site-packages (from transformers) (6.0)\n",
      "Requirement already satisfied: requests in /opt/conda/lib/python3.7/site-packages (from transformers) (2.28.1)\n",
      "Requirement already satisfied: importlib-metadata in /opt/conda/lib/python3.7/site-packages (from transformers) (4.13.0)\n",
      "Requirement already satisfied: regex!=2019.12.17 in /opt/conda/lib/python3.7/site-packages (from transformers) (2021.11.10)\n",
      "Requirement already satisfied: tokenizers!=0.11.3,<0.13,>=0.11.1 in /opt/conda/lib/python3.7/site-packages (from transformers) (0.12.1)\n",
      "Requirement already satisfied: packaging>=20.0 in /opt/conda/lib/python3.7/site-packages (from transformers) (21.3)\n",
      "Requirement already satisfied: filelock in /opt/conda/lib/python3.7/site-packages (from transformers) (3.7.1)\n",
      "Requirement already satisfied: huggingface-hub<1.0,>=0.1.0 in /opt/conda/lib/python3.7/site-packages (from transformers) (0.10.1)\n",
      "Requirement already satisfied: tqdm>=4.27 in /opt/conda/lib/python3.7/site-packages (from transformers) (4.64.0)\n",
      "Requirement already satisfied: numpy>=1.17 in /opt/conda/lib/python3.7/site-packages (from transformers) (1.21.6)\n",
      "Requirement already satisfied: typing-extensions>=3.7.4.3 in /opt/conda/lib/python3.7/site-packages (from huggingface-hub<1.0,>=0.1.0->transformers) (4.1.1)\n",
      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging>=20.0->transformers) (3.0.9)\n",
      "Requirement already satisfied: zipp>=0.5 in /opt/conda/lib/python3.7/site-packages (from importlib-metadata->transformers) (3.8.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /opt/conda/lib/python3.7/site-packages (from requests->transformers) (2022.9.24)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/lib/python3.7/site-packages (from requests->transformers) (1.26.12)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /opt/conda/lib/python3.7/site-packages (from requests->transformers) (3.3)\n",
      "Requirement already satisfied: charset-normalizer<3,>=2 in /opt/conda/lib/python3.7/site-packages (from requests->transformers) (2.1.0)\n",
      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "!pip install transformers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:10.913786Z",
     "iopub.status.busy": "2022-12-13T18:01:10.913389Z",
     "iopub.status.idle": "2022-12-13T18:01:10.940049Z",
     "shell.execute_reply": "2022-12-13T18:01:10.937792Z",
     "shell.execute_reply.started": "2022-12-13T18:01:10.913752Z"
    }
   },
   "outputs": [],
   "source": [
    "from helper_prabowo_ml import clean_html, remove_links, non_ascii, lower, email_address, removeStopWords, punct, remove_, remove_special_characters, remove_digits\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings, re\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from sklearn.model_selection import train_test_split\n",
    "from transformers import AutoTokenizer, TFBertModel\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.layers import Dense, Input, GlobalMaxPool1D, Dropout\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.initializers import TruncatedNormal\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.losses import CategoricalCrossentropy\n",
    "from tensorflow.keras.metrics import CategoricalAccuracy\n",
    "from sklearn.metrics import classification_report\n",
    "from tensorflow.keras.utils import plot_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the train, validation and test datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:10.951328Z",
     "iopub.status.busy": "2022-12-13T18:01:10.948594Z",
     "iopub.status.idle": "2022-12-13T18:01:11.034519Z",
     "shell.execute_reply": "2022-12-13T18:01:11.032823Z",
     "shell.execute_reply.started": "2022-12-13T18:01:10.951294Z"
    }
   },
   "outputs": [],
   "source": [
    "train = pd.read_csv(\"/kaggle/input/emotions-dataset-for-nlp/train.txt\",names=['Input','Sentiment'],sep=';',encoding='utf-8')\n",
    "val = pd.read_csv(\"/kaggle/input/emotions-dataset-for-nlp/val.txt\",names=['Input','Sentiment'],sep=';',encoding='utf-8')\n",
    "test = pd.read_csv(\"/kaggle/input/emotions-dataset-for-nlp/test.txt\",names=['Input','Sentiment'],sep=';',encoding='utf-8')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Combining the train, validation and test datasets into a single dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.041730Z",
     "iopub.status.busy": "2022-12-13T18:01:11.039476Z",
     "iopub.status.idle": "2022-12-13T18:01:11.063212Z",
     "shell.execute_reply": "2022-12-13T18:01:11.062361Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.041693Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Input</th>\n",
       "      <th>Sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>110</td>\n",
       "      <td>i feel it rarely advances any worthwhile cause...</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>14930</td>\n",
       "      <td>i have days weeks when i feel a little deprived</td>\n",
       "      <td>sadness</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>871</td>\n",
       "      <td>i feel suffocated and paranoid</td>\n",
       "      <td>fear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11109</td>\n",
       "      <td>i feel badly that my ability to be thrilled at...</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9070</td>\n",
       "      <td>i feel that your advice is only useful for the...</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index                                              Input Sentiment\n",
       "0    110  i feel it rarely advances any worthwhile cause...       joy\n",
       "1  14930    i have days weeks when i feel a little deprived   sadness\n",
       "2    871                     i feel suffocated and paranoid      fear\n",
       "3  11109  i feel badly that my ability to be thrilled at...       joy\n",
       "4   9070  i feel that your advice is only useful for the...       joy"
      ]
     },
     "execution_count": 259,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.concat([train,val,test],axis=0)\n",
    "df = df.sample(frac=0.1)\n",
    "df = df.reset_index()\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.069277Z",
     "iopub.status.busy": "2022-12-13T18:01:11.067107Z",
     "iopub.status.idle": "2022-12-13T18:01:11.076302Z",
     "shell.execute_reply": "2022-12-13T18:01:11.075169Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.069243Z"
    }
   },
   "outputs": [],
   "source": [
    "df.drop('index',axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.078381Z",
     "iopub.status.busy": "2022-12-13T18:01:11.077649Z",
     "iopub.status.idle": "2022-12-13T18:01:11.088774Z",
     "shell.execute_reply": "2022-12-13T18:01:11.087760Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.078346Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000, 2)"
      ]
     },
     "execution_count": 261,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.091482Z",
     "iopub.status.busy": "2022-12-13T18:01:11.090686Z",
     "iopub.status.idle": "2022-12-13T18:01:11.106464Z",
     "shell.execute_reply": "2022-12-13T18:01:11.105520Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.091440Z"
    }
   },
   "outputs": [],
   "source": [
    "def preprocess_data(data,col):\n",
    "    data[col] = data[col].apply(func=clean_html)\n",
    "    data[col] = data[col].apply(func=remove_digits)\n",
    "    data[col] = data[col].apply(func=remove_)\n",
    "    data[col] = data[col].apply(func=removeStopWords)\n",
    "    data[col] = data[col].apply(func=remove_links)\n",
    "    data[col] = data[col].apply(func=remove_special_characters)\n",
    "    data[col] = data[col].apply(func=non_ascii)\n",
    "    data[col] = data[col].apply(func=email_address)\n",
    "    data[col] = data[col].apply(func=punct)\n",
    "    data[col] = data[col].apply(func=lower)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.108375Z",
     "iopub.status.busy": "2022-12-13T18:01:11.107777Z",
     "iopub.status.idle": "2022-12-13T18:01:11.604565Z",
     "shell.execute_reply": "2022-12-13T18:01:11.603545Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.108342Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Input</th>\n",
       "      <th>Sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>feel rarely advances worthwhile cause always s...</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>days weeks feel little deprived</td>\n",
       "      <td>sadness</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>feel suffocated paranoid</td>\n",
       "      <td>fear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>feel badly ability thrilled seeing something p...</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>feel advice useful people already believe</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               Input Sentiment\n",
       "0  feel rarely advances worthwhile cause always s...       joy\n",
       "1                    days weeks feel little deprived   sadness\n",
       "2                           feel suffocated paranoid      fear\n",
       "3  feel badly ability thrilled seeing something p...       joy\n",
       "4          feel advice useful people already believe       joy"
      ]
     },
     "execution_count": 263,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocessed_df = preprocess_data(df,'Input')\n",
    "preprocessed_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.609507Z",
     "iopub.status.busy": "2022-12-13T18:01:11.609193Z",
     "iopub.status.idle": "2022-12-13T18:01:11.615773Z",
     "shell.execute_reply": "2022-12-13T18:01:11.614790Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.609480Z"
    }
   },
   "outputs": [],
   "source": [
    "preprocessed_df['num_words'] = preprocessed_df.Input.apply(len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.617661Z",
     "iopub.status.busy": "2022-12-13T18:01:11.617122Z",
     "iopub.status.idle": "2022-12-13T18:01:11.632703Z",
     "shell.execute_reply": "2022-12-13T18:01:11.631224Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.617626Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Input</th>\n",
       "      <th>Sentiment</th>\n",
       "      <th>num_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>feel rarely advances worthwhile cause always s...</td>\n",
       "      <td>joy</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>days weeks feel little deprived</td>\n",
       "      <td>sadness</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>feel suffocated paranoid</td>\n",
       "      <td>fear</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>feel badly ability thrilled seeing something p...</td>\n",
       "      <td>joy</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>feel advice useful people already believe</td>\n",
       "      <td>joy</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               Input Sentiment  num_words\n",
       "0  feel rarely advances worthwhile cause always s...       joy        115\n",
       "1                    days weeks feel little deprived   sadness         31\n",
       "2                           feel suffocated paranoid      fear         24\n",
       "3  feel badly ability thrilled seeing something p...       joy         57\n",
       "4          feel advice useful people already believe       joy         41"
      ]
     },
     "execution_count": 265,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocessed_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.634854Z",
     "iopub.status.busy": "2022-12-13T18:01:11.634594Z",
     "iopub.status.idle": "2022-12-13T18:01:11.643710Z",
     "shell.execute_reply": "2022-12-13T18:01:11.642615Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.634830Z"
    }
   },
   "outputs": [],
   "source": [
    "encoded_labels = {'anger': 0, 'fear': 1, 'joy': 2, 'love': 3, 'sadness': 4, 'surprise': 5} "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train-Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.645388Z",
     "iopub.status.busy": "2022-12-13T18:01:11.644880Z",
     "iopub.status.idle": "2022-12-13T18:01:11.658020Z",
     "shell.execute_reply": "2022-12-13T18:01:11.656969Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.645354Z"
    }
   },
   "outputs": [],
   "source": [
    "train_data, test_data = train_test_split(preprocessed_df,test_size=0.3,random_state=101,shuffle=True,stratify=preprocessed_df.Sentiment)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the Tokenizer class and pretrained BERT model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:11.660880Z",
     "iopub.status.busy": "2022-12-13T18:01:11.660056Z",
     "iopub.status.idle": "2022-12-13T18:01:14.660889Z",
     "shell.execute_reply": "2022-12-13T18:01:14.659826Z",
     "shell.execute_reply.started": "2022-12-13T18:01:11.660841Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some layers from the model checkpoint at bert-base-uncased were not used when initializing TFBertModel: ['nsp___cls', 'mlm___cls']\n",
      "- This IS expected if you are initializing TFBertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing TFBertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "All the layers of TFBertModel were initialized from the model checkpoint at bert-base-uncased.\n",
      "If your task is similar to the task the model of the checkpoint was trained on, you can already use TFBertModel for predictions without further training.\n"
     ]
    }
   ],
   "source": [
    "tokenizer = AutoTokenizer.from_pretrained(\"bert-base-uncased\")\n",
    "bert_model = TFBertModel.from_pretrained(\"bert-base-uncased\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:14.664379Z",
     "iopub.status.busy": "2022-12-13T18:01:14.663276Z",
     "iopub.status.idle": "2022-12-13T18:01:15.027035Z",
     "shell.execute_reply": "2022-12-13T18:01:15.026061Z",
     "shell.execute_reply.started": "2022-12-13T18:01:14.664337Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fe278d08990>"
      ]
     },
     "execution_count": 269,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(preprocessed_df.num_words,height=8,aspect=1.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:15.030135Z",
     "iopub.status.busy": "2022-12-13T18:01:15.029154Z",
     "iopub.status.idle": "2022-12-13T18:01:15.034593Z",
     "shell.execute_reply": "2022-12-13T18:01:15.033530Z",
     "shell.execute_reply.started": "2022-12-13T18:01:15.030088Z"
    }
   },
   "outputs": [],
   "source": [
    "max_len = 40"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text Tokenization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:15.036578Z",
     "iopub.status.busy": "2022-12-13T18:01:15.036031Z",
     "iopub.status.idle": "2022-12-13T18:01:15.140999Z",
     "shell.execute_reply": "2022-12-13T18:01:15.140145Z",
     "shell.execute_reply.started": "2022-12-13T18:01:15.036543Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train = tokenizer(text=train_data.Input.tolist(),\n",
    "                   add_special_tokens=True,\n",
    "                   return_tensors='tf',\n",
    "                   max_length=max_len,\n",
    "                   padding=True,\n",
    "                   truncation=True,\n",
    "                   return_token_type_ids=False,\n",
    "                   return_attention_mask=True,\n",
    "                   verbose=True\n",
    "                   )\n",
    "\n",
    "X_test = tokenizer(text=test_data.Input.tolist(),\n",
    "                   add_special_tokens=True,\n",
    "                   return_tensors='tf',\n",
    "                   max_length=max_len,\n",
    "                   padding=True,\n",
    "                   truncation=True,\n",
    "                   return_token_type_ids=False,\n",
    "                   return_attention_mask=True,\n",
    "                   verbose=True\n",
    "                  )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Defining the model architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:15.142568Z",
     "iopub.status.busy": "2022-12-13T18:01:15.142246Z",
     "iopub.status.idle": "2022-12-13T18:01:15.151049Z",
     "shell.execute_reply": "2022-12-13T18:01:15.150081Z",
     "shell.execute_reply.started": "2022-12-13T18:01:15.142533Z"
    }
   },
   "outputs": [],
   "source": [
    "input_ids = Input(shape=(max_len,),name='input_ids',dtype=tf.int32)\n",
    "attention_mask = Input(shape=(max_len,),name='attention_mask',dtype=tf.int32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:15.153257Z",
     "iopub.status.busy": "2022-12-13T18:01:15.152619Z",
     "iopub.status.idle": "2022-12-13T18:01:16.727568Z",
     "shell.execute_reply": "2022-12-13T18:01:16.726591Z",
     "shell.execute_reply.started": "2022-12-13T18:01:15.153223Z"
    }
   },
   "outputs": [],
   "source": [
    "embeddings = bert_model(input_ids,attention_mask=attention_mask)[0] # 0: final hidden state, 1: pooling output\n",
    "output = GlobalMaxPool1D()(embeddings)\n",
    "output = Dense(units=128,activation='relu')(output)\n",
    "output = Dropout(0.1)(output)\n",
    "output = Dense(units=64,activation='relu')(output)\n",
    "output = Dense(units=32,activation='relu')(output)\n",
    "y = Dense(units=6,activation='softmax')(output)\n",
    "\n",
    "model = Model(inputs=[input_ids,attention_mask],outputs=y)\n",
    "model.layers[2].trainable = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compiling the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.729456Z",
     "iopub.status.busy": "2022-12-13T18:01:16.729060Z",
     "iopub.status.idle": "2022-12-13T18:01:16.747385Z",
     "shell.execute_reply": "2022-12-13T18:01:16.746538Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.729421Z"
    }
   },
   "outputs": [],
   "source": [
    "model.compile(loss=CategoricalCrossentropy(from_logits=True),\n",
    "              optimizer=Adam(learning_rate=5e-5,epsilon=1e-8,decay=0.01,clipnorm=1.0),\n",
    "              metrics=CategoricalAccuracy('balanced_accuracy'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Encoding the emotion labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.749068Z",
     "iopub.status.busy": "2022-12-13T18:01:16.748630Z",
     "iopub.status.idle": "2022-12-13T18:01:16.758232Z",
     "shell.execute_reply": "2022-12-13T18:01:16.757304Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.749021Z"
    }
   },
   "outputs": [],
   "source": [
    "train_data['Label'] = train_data.Sentiment.map(encoded_labels)\n",
    "test_data['Label'] = test_data.Sentiment.map(encoded_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.760086Z",
     "iopub.status.busy": "2022-12-13T18:01:16.759642Z",
     "iopub.status.idle": "2022-12-13T18:01:16.771817Z",
     "shell.execute_reply": "2022-12-13T18:01:16.770678Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.760050Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Input</th>\n",
       "      <th>Sentiment</th>\n",
       "      <th>num_words</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1492</th>\n",
       "      <td>hurt feelings sorry</td>\n",
       "      <td>sadness</td>\n",
       "      <td>19</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>feel kinda weird andrea tries talk chris</td>\n",
       "      <td>fear</td>\n",
       "      <td>40</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1372</th>\n",
       "      <td>feel ungrateful unhappy seem move properly</td>\n",
       "      <td>sadness</td>\n",
       "      <td>42</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>433</th>\n",
       "      <td>feel kids troubled homes feel ones could chang...</td>\n",
       "      <td>sadness</td>\n",
       "      <td>54</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>943</th>\n",
       "      <td>feel angry feel sad</td>\n",
       "      <td>anger</td>\n",
       "      <td>19</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  Input Sentiment  num_words  \\\n",
       "1492                                hurt feelings sorry   sadness         19   \n",
       "162            feel kinda weird andrea tries talk chris      fear         40   \n",
       "1372         feel ungrateful unhappy seem move properly   sadness         42   \n",
       "433   feel kids troubled homes feel ones could chang...   sadness         54   \n",
       "943                                 feel angry feel sad     anger         19   \n",
       "\n",
       "      Label  \n",
       "1492      4  \n",
       "162       1  \n",
       "1372      4  \n",
       "433       4  \n",
       "943       0  "
      ]
     },
     "execution_count": 276,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.773868Z",
     "iopub.status.busy": "2022-12-13T18:01:16.773519Z",
     "iopub.status.idle": "2022-12-13T18:01:16.786695Z",
     "shell.execute_reply": "2022-12-13T18:01:16.785643Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.773834Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Input</th>\n",
       "      <th>Sentiment</th>\n",
       "      <th>num_words</th>\n",
       "      <th>Label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1445</th>\n",
       "      <td>basically wrote future self feeling generous t...</td>\n",
       "      <td>love</td>\n",
       "      <td>78</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>feel safer didnt horrible panicky feeling used...</td>\n",
       "      <td>sadness</td>\n",
       "      <td>90</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>670</th>\n",
       "      <td>cluded feelings violent acts committed</td>\n",
       "      <td>anger</td>\n",
       "      <td>38</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>581</th>\n",
       "      <td>feel hostile toward others fail nice</td>\n",
       "      <td>anger</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1117</th>\n",
       "      <td>feeling invigorated ready keep pushing goa</td>\n",
       "      <td>joy</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  Input Sentiment  num_words  \\\n",
       "1445  basically wrote future self feeling generous t...      love         78   \n",
       "51    feel safer didnt horrible panicky feeling used...   sadness         90   \n",
       "670              cluded feelings violent acts committed     anger         38   \n",
       "581                feel hostile toward others fail nice     anger         36   \n",
       "1117         feeling invigorated ready keep pushing goa       joy         42   \n",
       "\n",
       "      Label  \n",
       "1445      3  \n",
       "51        4  \n",
       "670       0  \n",
       "581       0  \n",
       "1117      2  "
      ]
     },
     "execution_count": 277,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Generating the model summary and plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.788461Z",
     "iopub.status.busy": "2022-12-13T18:01:16.787996Z",
     "iopub.status.idle": "2022-12-13T18:01:16.809105Z",
     "shell.execute_reply": "2022-12-13T18:01:16.807712Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.788427Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_9\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_ids (InputLayer)          [(None, 40)]         0                                            \n",
      "__________________________________________________________________________________________________\n",
      "attention_mask (InputLayer)     [(None, 40)]         0                                            \n",
      "__________________________________________________________________________________________________\n",
      "tf_bert_model_9 (TFBertModel)   TFBaseModelOutputWit 109482240   input_ids[0][0]                  \n",
      "                                                                 attention_mask[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "global_max_pooling1d_9 (GlobalM (None, 768)          0           tf_bert_model_9[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "dense_27 (Dense)                (None, 128)          98432       global_max_pooling1d_9[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dropout_379 (Dropout)           (None, 128)          0           dense_27[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_28 (Dense)                (None, 64)           8256        dropout_379[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "dense_29 (Dense)                (None, 32)           2080        dense_28[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_30 (Dense)                (None, 6)            198         dense_29[0][0]                   \n",
      "==================================================================================================\n",
      "Total params: 109,591,206\n",
      "Trainable params: 109,591,206\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:16.824422Z",
     "iopub.status.busy": "2022-12-13T18:01:16.823769Z",
     "iopub.status.idle": "2022-12-13T18:01:17.507350Z",
     "shell.execute_reply": "2022-12-13T18:01:17.506142Z",
     "shell.execute_reply.started": "2022-12-13T18:01:16.824385Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
      "To disable this warning, you can either:\n",
      "\t- Avoid using `tokenizers` before the fork if possible\n",
      "\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
      "huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
      "To disable this warning, you can either:\n",
      "\t- Avoid using `tokenizers` before the fork if possible\n",
      "\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(model,'model.png',show_shapes=True,dpi=100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training and fine-tuning the pretrained BERT model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:01:17.510944Z",
     "iopub.status.busy": "2022-12-13T18:01:17.509027Z",
     "iopub.status.idle": "2022-12-13T18:05:01.087510Z",
     "shell.execute_reply": "2022-12-13T18:05:01.086444Z",
     "shell.execute_reply.started": "2022-12-13T18:01:17.510897Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "44/44 [==============================] - 40s 413ms/step - loss: 1.6248 - balanced_accuracy: 0.3993 - val_loss: 1.3380 - val_balanced_accuracy: 0.5500\n",
      "Epoch 2/10\n",
      "44/44 [==============================] - 14s 312ms/step - loss: 1.1277 - balanced_accuracy: 0.6336 - val_loss: 0.8645 - val_balanced_accuracy: 0.7217\n",
      "Epoch 3/10\n",
      "44/44 [==============================] - 14s 321ms/step - loss: 0.6700 - balanced_accuracy: 0.8057 - val_loss: 0.5863 - val_balanced_accuracy: 0.8083\n",
      "Epoch 4/10\n",
      "44/44 [==============================] - 14s 329ms/step - loss: 0.3557 - balanced_accuracy: 0.8943 - val_loss: 0.4998 - val_balanced_accuracy: 0.8433\n",
      "Epoch 5/10\n",
      "44/44 [==============================] - 15s 331ms/step - loss: 0.2323 - balanced_accuracy: 0.9457 - val_loss: 0.4373 - val_balanced_accuracy: 0.8733\n",
      "Epoch 6/10\n",
      "44/44 [==============================] - 14s 323ms/step - loss: 0.1614 - balanced_accuracy: 0.9607 - val_loss: 0.4115 - val_balanced_accuracy: 0.8767\n",
      "Epoch 7/10\n",
      "44/44 [==============================] - 14s 320ms/step - loss: 0.0977 - balanced_accuracy: 0.9821 - val_loss: 0.4457 - val_balanced_accuracy: 0.8750\n",
      "Epoch 8/10\n",
      "44/44 [==============================] - 14s 322ms/step - loss: 0.0669 - balanced_accuracy: 0.9886 - val_loss: 0.3988 - val_balanced_accuracy: 0.8900\n",
      "Epoch 9/10\n",
      "44/44 [==============================] - 14s 320ms/step - loss: 0.0491 - balanced_accuracy: 0.9971 - val_loss: 0.4291 - val_balanced_accuracy: 0.8850\n",
      "Epoch 10/10\n",
      "44/44 [==============================] - 14s 324ms/step - loss: 0.0366 - balanced_accuracy: 0.9979 - val_loss: 0.4431 - val_balanced_accuracy: 0.8833\n"
     ]
    }
   ],
   "source": [
    "r = model.fit(x={'input_ids': X_train['input_ids'], 'attention_mask': X_train['attention_mask']},\n",
    "             y=to_categorical(train_data.Label),\n",
    "             epochs=10,\n",
    "             batch_size=32,\n",
    "             validation_data=({'input_ids': X_test['input_ids'], 'attention_mask': X_test['attention_mask']},to_categorical(test_data.Label))\n",
    "             )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analyzing model performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:05:01.090755Z",
     "iopub.status.busy": "2022-12-13T18:05:01.090348Z",
     "iopub.status.idle": "2022-12-13T18:05:01.341971Z",
     "shell.execute_reply": "2022-12-13T18:05:01.340961Z",
     "shell.execute_reply.started": "2022-12-13T18:05:01.090718Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,8))\n",
    "plt.plot(r.history['loss'],'r',label='train loss')\n",
    "plt.plot(r.history['val_loss'],'b',label='test loss')\n",
    "plt.xlabel('No. of Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.title('Loss Graph')\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:05:01.347917Z",
     "iopub.status.busy": "2022-12-13T18:05:01.347043Z",
     "iopub.status.idle": "2022-12-13T18:05:01.587962Z",
     "shell.execute_reply": "2022-12-13T18:05:01.587052Z",
     "shell.execute_reply.started": "2022-12-13T18:05:01.347880Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,8))\n",
    "plt.plot(r.history['balanced_accuracy'],'r',label='train accuracy')\n",
    "plt.plot(r.history['val_balanced_accuracy'],'b',label='test accuracy')\n",
    "plt.xlabel('No. of Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.title('Accuracy Graph')\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Saving the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:05:18.984902Z",
     "iopub.status.busy": "2022-12-13T18:05:18.984526Z",
     "iopub.status.idle": "2022-12-13T18:05:22.792132Z",
     "shell.execute_reply": "2022-12-13T18:05:22.790967Z",
     "shell.execute_reply.started": "2022-12-13T18:05:18.984872Z"
    }
   },
   "outputs": [],
   "source": [
    "model.save(\"emotion_detector.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating the model on the test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:05:22.795044Z",
     "iopub.status.busy": "2022-12-13T18:05:22.794443Z",
     "iopub.status.idle": "2022-12-13T18:05:28.759338Z",
     "shell.execute_reply": "2022-12-13T18:05:28.758346Z",
     "shell.execute_reply.started": "2022-12-13T18:05:22.795003Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19/19 [==============================] - 5s 91ms/step - loss: 1.1729 - balanced_accuracy: 0.8833\n",
      "Test Categorical Cross-Entropy Loss: 1.1729487180709839\n",
      "Test Categorical Accuracy: 0.8833333253860474\n"
     ]
    }
   ],
   "source": [
    "loss, acc = model.evaluate({'input_ids': X_test['input_ids'], 'attention_mask': X_test['attention_mask']},to_categorical(test_data.Label))\n",
    "print(\"Test Categorical Cross-Entropy Loss:\",loss)\n",
    "print(\"Test Categorical Accuracy:\",acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-13T18:05:28.761706Z",
     "iopub.status.busy": "2022-12-13T18:05:28.761010Z",
     "iopub.status.idle": "2022-12-13T18:05:31.759064Z",
     "shell.execute_reply": "2022-12-13T18:05:31.757861Z",
     "shell.execute_reply.started": "2022-12-13T18:05:28.761668Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.90      0.89        77\n",
      "           1       0.82      0.85      0.83        68\n",
      "           2       0.93      0.89      0.91       209\n",
      "           3       0.69      0.81      0.75        54\n",
      "           4       0.92      0.92      0.92       173\n",
      "           5       0.88      0.74      0.80        19\n",
      "\n",
      "    accuracy                           0.88       600\n",
      "   macro avg       0.85      0.85      0.85       600\n",
      "weighted avg       0.89      0.88      0.88       600\n",
      "\n"
     ]
    }
   ],
   "source": [
    "test_predictions = model.predict({'input_ids': X_test['input_ids'], 'attention_mask': X_test['attention_mask']})\n",
    "test_predictions = np.argmax(test_predictions,axis=1)\n",
    "print(classification_report(test_data.Label,test_predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The result is quite promising as we've obtained an excellent f1-score of almost 80% for all the 6 emotion classes."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}