{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Notebook Fine-Tuning Bert\n",
    "In diesem Notebook wird Bert bzw. 'BertForSequenceClassification' feingetuned. <br>\n",
    "Funktionen werden aus diesem [Skript](bert_no_ernie.py) geladen."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from bert_no_ernie import *\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rohdaten einlesen\n",
    "An dieser Stelle, wird der Hackathon Datensatz eingelesen welcher Annotierte Daten enthält.\n",
    "Die wichtigsten Attribute dieses Datensatzes in diesem sind *Text* (welcher den \"Witz\" als String enthält) und *is_humor* (ein durch 0 und 1 dargestellter Wahrheitswert) welcher angibt ob der entsprechende Text in der Zeile ein Witz ist oder nicht."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>id</th>\n",
       "      <th>text</th>\n",
       "      <th>is_humor</th>\n",
       "      <th>humor_rating</th>\n",
       "      <th>humor_controversy</th>\n",
       "      <th>offense_rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>TENNESSEE: We're the best state. Nobody even c...</td>\n",
       "      <td>1</td>\n",
       "      <td>2.42</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>A man inserted an advertisement in the classif...</td>\n",
       "      <td>1</td>\n",
       "      <td>2.50</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>How many men does it take to open a can of bee...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.95</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>Told my mom I hit 1200 Twitter followers. She ...</td>\n",
       "      <td>1</td>\n",
       "      <td>2.11</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>Roses are dead. Love is fake. Weddings are bas...</td>\n",
       "      <td>1</td>\n",
       "      <td>2.78</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id                                               text  is_humor  \\\n",
       "0   1  TENNESSEE: We're the best state. Nobody even c...         1   \n",
       "1   2  A man inserted an advertisement in the classif...         1   \n",
       "2   3  How many men does it take to open a can of bee...         1   \n",
       "3   4  Told my mom I hit 1200 Twitter followers. She ...         1   \n",
       "4   5  Roses are dead. Love is fake. Weddings are bas...         1   \n",
       "\n",
       "   humor_rating  humor_controversy  offense_rating  \n",
       "0          2.42                1.0             0.2  \n",
       "1          2.50                1.0             1.1  \n",
       "2          1.95                0.0             2.4  \n",
       "3          2.11                1.0             0.0  \n",
       "4          2.78                0.0             0.1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"data/hack.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Hyperparameter festlegen. Und Zufall seeden\n",
    "# Set Max Epoch Amount\n",
    "EPOCH = 10\n",
    "# DROPOUT-PROBABILITY\n",
    "DROPOUT = 0.1\n",
    "# BATCHSIZE\n",
    "BATCH_SIZE = 16\n",
    "#LEARNING RATE\n",
    "LEARNING_RATE = 1e-5\n",
    "# RANDOM SEED\n",
    "RNDM_SEED = 501\n",
    "# FREEZE Bert Layers\n",
    "FREEZE = True\n",
    "\n",
    "torch.manual_seed(RNDM_SEED)\n",
    "np.random.seed(RNDM_SEED)\n",
    "torch.cuda.manual_seed_all(RNDM_SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tokenizer für Bert Model laden.\n",
    "tokenizer = AutoTokenizer.from_pretrained(\"google-bert/bert-base-uncased\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Daten aufteilen(70/15/15) und an Custom Dataset Klasse übergeben\n",
    "train_data,test_data,val_data = create_datasets(tokenizer,df,.7,True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DataLoaders basierend auf Datasets kreieren.\n",
    "train_loader, test_loader, validation_loader = create_dataloaders([train_data,test_data,val_data],batchsize=BATCH_SIZE,shufflelist=[True,True,False])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of BertForSequenceClassification were not initialized from the model checkpoint at bert-base-uncased and are newly initialized: ['classifier.bias', 'classifier.weight']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
     ]
    }
   ],
   "source": [
    "# Model instanziieren, sowie Loss-Funktion und Optimizer\n",
    "mybert = CustomBert(DROPOUT)\n",
    "mybert.to(DEVICE)\n",
    "\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.Adam(mybert.parameters(), lr = LEARNING_RATE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For 1 the Scores are: \n",
      "Training Loss is 0.6827\n",
      "Validation Loss: 0.6828 ### Validation Accuracy 60.8333%\n",
      "For 2 the Scores are: \n",
      "Training Loss is 0.6836\n",
      "Validation Loss: 0.6825 ### Validation Accuracy 60.8333%\n",
      "For 3 the Scores are: \n",
      "Training Loss is 0.6824\n",
      "Validation Loss: 0.6821 ### Validation Accuracy 60.8333%\n",
      "For 4 the Scores are: \n",
      "Training Loss is 0.6815\n",
      "Validation Loss: 0.6817 ### Validation Accuracy 60.8333%\n",
      "For 5 the Scores are: \n",
      "Training Loss is 0.6808\n",
      "Validation Loss: 0.6814 ### Validation Accuracy 60.8333%\n",
      "For 6 the Scores are: \n",
      "Training Loss is 0.6809\n",
      "Validation Loss: 0.6810 ### Validation Accuracy 60.8333%\n",
      "For 7 the Scores are: \n",
      "Training Loss is 0.6801\n",
      "Validation Loss: 0.6807 ### Validation Accuracy 60.7500%\n",
      "For 8 the Scores are: \n",
      "Training Loss is 0.6795\n",
      "Validation Loss: 0.6804 ### Validation Accuracy 60.7500%\n",
      "For 9 the Scores are: \n",
      "Training Loss is 0.6797\n",
      "Validation Loss: 0.6801 ### Validation Accuracy 60.7500%\n",
      "For 10 the Scores are: \n",
      "Training Loss is 0.6793\n",
      "Validation Loss: 0.6799 ### Validation Accuracy 60.7500%\n"
     ]
    }
   ],
   "source": [
    "# Trainings - und Validierungs Durchgänge\n",
    "loss_vals, eval_vals = np.zeros(EPOCH), np.zeros(EPOCH)\n",
    "\n",
    "for epoch in range(EPOCH):\n",
    "    print(f\"For {epoch+1} the Scores are: \")\n",
    "    loss_vals[epoch] = training_loop(mybert,optimizer=optimizer,criterion=criterion,train_loader=train_loader,freeze_bert=FREEZE)\n",
    "    eval_vals[epoch] = eval_loop(mybert,criterion=criterion,validation_loader=validation_loader)    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.68267711, 0.68355761, 0.68237029, 0.68148399, 0.68079539,\n",
       "       0.68086683, 0.68012043, 0.67948493, 0.67972843, 0.67932365])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([0.68283186, 0.68245001, 0.68208028, 0.68170239, 0.68136094,\n",
       "       0.68103237, 0.68071597, 0.68041458, 0.68011246, 0.67985092])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(loss_vals)\n",
    "display(eval_vals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test_loop(model:CustomBert, test_loader:DataLoader):\n",
    "    test_accuracy = np.zeros(len(test_loader))\n",
    "    for index,batch in enumerate(test_loader):\n",
    "        input_ids, att_mask, labels = batch.values()\n",
    "        input_ids, att_mask, labels = input_ids.to(DEVICE), att_mask.to(DEVICE), labels.to(DEVICE)\n",
    "        with torch.no_grad():\n",
    "            # model = torch.load(\"best_bert_model.pth\")\n",
    "            # model.to(DEVICE)\n",
    "            output = model(input_ids,att_mask)\n",
    "            output = output.cpu()\n",
    "            labels = labels.cpu()\n",
    "            pred_flat = np.argmax(a=output,axis=1).flatten()\n",
    "            test_accuracy[index] = accuracy_score(labels,pred_flat)\n",
    "\n",
    "    return test_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.6875, 0.5625, 0.75  , 0.625 , 0.625 , 0.75  , 0.6875, 0.5   ,\n",
       "       0.375 , 0.1875, 0.4375, 0.75  , 0.75  , 0.8125, 0.5   , 0.5   ,\n",
       "       0.8125, 0.5   , 0.8125, 0.625 , 0.5625, 0.4375, 0.5625, 0.8125,\n",
       "       0.6875, 0.8125, 0.625 , 0.6875, 0.5625, 0.75  , 0.8125, 0.8125,\n",
       "       0.75  , 0.5   , 0.625 , 0.6875, 0.6875, 0.5   , 0.625 , 0.5625,\n",
       "       0.625 , 0.4375, 0.6875, 0.75  , 0.6875, 0.1875, 0.625 , 0.5   ,\n",
       "       0.875 , 0.625 , 0.625 , 0.4375, 0.5625, 0.6875, 0.6875, 0.625 ,\n",
       "       0.375 , 0.4375, 0.6875, 0.6875, 0.5625, 0.4375, 0.5   , 0.5625,\n",
       "       0.6875, 0.5625, 0.4375, 0.8125, 0.75  , 0.75  , 0.625 , 0.6875,\n",
       "       0.5625, 0.9375, 0.5625])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_acc_score = test_loop(mybert,test_loader)\n",
    "test_acc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_test_metrics(accuracy):\n",
    "    \"\"\"\n",
    "    Plot Test Metrics of Model (Confiuson Matrix, Accuracy)\n",
    "    \"\"\"\n",
    "    plt.plot(accuracy)\n",
    "    plt.hlines(np.mean(accuracy),0,len(accuracy),'red','dotted','Mean Accuracy {:.4f}'.format(np.mean(accuracy)))\n",
    "    plt.title(\"Accuracy of Test\")\n",
    "    plt.xlabel(\"Num Batches\")\n",
    "    plt.ylabel(\"Accurcy 0.0 - 1.0\")\n",
    "    plt.grid(True)\n",
    "    plt.legend()\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_test_metrics(test_acc_score)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}