Sentiment Analysis/Financial Sentiment Analysis/Financial_Sentiment_Analysis_v2.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sentiment Analysis of Statements about Finances\n",
    "\n",
    "Source for the Datasets: \n",
    "- https://www.kaggle.com/datasets/sbhatti/financial-sentiment-analysis\n",
    "- https://www.kaggle.com/datasets/ankurzing/sentiment-analysis-for-financial-news"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exploratory Data Analysis & Preprocessing in Pandas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing the Necessary Libraries for the Entire Project"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Git LFS initialized.\n"
     ]
    }
   ],
   "source": [
    "import os, sys\n",
    "os.environ['TOKENIZERS_PARALLELISM']='false'\n",
    "\n",
    "from pathlib import Path\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import torch\n",
    "from torch import nn\n",
    "\n",
    "import datasets\n",
    "from datasets import Dataset, DatasetDict\n",
    "\n",
    "import transformers\n",
    "from transformers import AutoModelForSequenceClassification, AutoTokenizer\n",
    "from transformers import TrainingArguments, Trainer, set_seed, DataCollatorWithPadding\n",
    "\n",
    "import evaluate\n",
    "\n",
    "!git lfs install"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Versions of Libraries Used"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "           Python: 3.9.12\n",
      "            NumPy: 1.23.5\n",
      "           Pandas: 2.0.0\n",
      "         Datasets: 2.11.0\n",
      "     Transformers: 4.27.4\n",
      "            Torch: 2.0.0\n",
      "         Evaluate: 0.4.0\n"
     ]
    }
   ],
   "source": [
    "n = 18\n",
    "\n",
    "print(\"Python:\".rjust(n), sys.version[0:6])\n",
    "print(\"NumPy:\".rjust(n), np.__version__)\n",
    "print(\"Pandas:\".rjust(n), pd.__version__)\n",
    "print(\"Datasets:\".rjust(n), datasets.__version__)\n",
    "print(\"Transformers:\".rjust(n), transformers.__version__)\n",
    "print(\"Torch:\".rjust(n), torch.__version__)\n",
    "print(\"Evaluate:\".rjust(n), evaluate.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ingest First Dataset & Preprocess it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 5836 entries, 0 to 5841\n",
      "Data columns (total 2 columns):\n",
      " #   Column  Non-Null Count  Dtype \n",
      "---  ------  --------------  ----- \n",
      " 0   text    5836 non-null   object\n",
      " 1   label   5836 non-null   object\n",
      "dtypes: object(2)\n",
      "memory usage: 136.8+ KB\n"
     ]
    }
   ],
   "source": [
    "data1 = pd.read_csv(\"data.csv\", \n",
    "                    engine=\"c\",\n",
    "                    names=[\"text\",\n",
    "                           \"label\"],\n",
    "                    header=0\n",
    "                    )\n",
    "\n",
    "data1 = data1.dropna()\n",
    "data1 = data1.drop_duplicates()\n",
    "\n",
    "data1['text'] = data1['text'].str.strip()\n",
    "data1['label'] = data1['label'].str.strip()\n",
    "\n",
    "data1.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ingest Second Dataset & Preprocess it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 4839 entries, 0 to 4844\n",
      "Data columns (total 2 columns):\n",
      " #   Column  Non-Null Count  Dtype \n",
      "---  ------  --------------  ----- \n",
      " 0   label   4839 non-null   object\n",
      " 1   text    4839 non-null   object\n",
      "dtypes: object(2)\n",
      "memory usage: 113.4+ KB\n"
     ]
    }
   ],
   "source": [
    "data2 = pd.read_csv(\"all-data.csv\", \n",
    "                    encoding_errors='ignore', \n",
    "                    engine=\"c\", \n",
    "                    names=['label', \n",
    "                           'text'],\n",
    "                    header=0\n",
    "                    )\n",
    "\n",
    "data2['label'] = data2['label'].str.strip()\n",
    "data2['text'] = data2['text'].str.strip()\n",
    "\n",
    "data2 = data2.drop_duplicates()\n",
    "data2.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Combine Both Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>text</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>The GeoSolutions technology will leverage Bene...</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>$ESI on lows, down $1.50 to $2.50 BK a real po...</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>For the last quarter of 2010 , Componenta 's n...</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>According to the Finnish-Russian Chamber of Co...</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>The Swedish buyout firm has sold its remaining...</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text     label\n",
       "0  The GeoSolutions technology will leverage Bene...  positive\n",
       "1  $ESI on lows, down $1.50 to $2.50 BK a real po...  negative\n",
       "2  For the last quarter of 2010 , Componenta 's n...  positive\n",
       "3  According to the Finnish-Russian Chamber of Co...   neutral\n",
       "4  The Swedish buyout firm has sold its remaining...   neutral"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.concat([data1, data2]).reset_index(drop=True)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Histogram of Text Lengths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The maximum length of Speeches is 81 words.\n"
     ]
    }
   ],
   "source": [
    "data['text_length'] = data['text'].str.split().str.len()\n",
    "\n",
    "n_bin=82\n",
    "data.hist('text_length', \n",
    "          grid=False, \n",
    "          figsize=(18,8), \n",
    "          range=(0,n_bin), \n",
    "          bins=n_bin)\n",
    "plt.title('Histogram of Tweet Lengths')\n",
    "plt.xlabel('# of Words in Tweets')\n",
    "plt.ylabel('# of Samples')\n",
    "plt.show()\n",
    "\n",
    "print(f\"The maximum length of Speeches is {max(data['text_length'])} words.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Visualize: Description Lengths By Output/Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.boxplot(\"text_length\", \n",
    "             by=\"label\", \n",
    "             grid=True, \n",
    "             showfliers=True,\n",
    "             figsize=(10,4))\n",
    "\n",
    "plt.suptitle('')\n",
    "plt.title('Review Length by Type')\n",
    "plt.xlabel('Sentiment Class')\n",
    "plt.ylabel('Words per Review')\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Class Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "class_counts = data.groupby(\"label\").count()\n",
    "\n",
    "class_counts['text_length'].plot.bar(label='labels', \n",
    "                                     color=plt.cm.Paired(np.arange(len(data))))\n",
    "plt.suptitle('')\n",
    "plt.title('Class Distribution')\n",
    "plt.xlabel('Class')\n",
    "plt.ylabel('# of Samples')\n",
    "plt.xticks(rotation = 0)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Dictionaries to Convert Between String & Integer Class Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "unique_classes = data['label'].unique()\n",
    "\n",
    "NUM_OF_LABELS = len(unique_classes)\n",
    "\n",
    "id2label = {idx:label for idx, label in enumerate(unique_classes)}\n",
    "label2id = {label:idx for idx, label in enumerate(unique_classes)}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Remove Overly Lengthy Samples & Unnecessary Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>text</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>The GeoSolutions technology will leverage Bene...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>$ESI on lows, down $1.50 to $2.50 BK a real po...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>For the last quarter of 2010 , Componenta 's n...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>According to the Finnish-Russian Chamber of Co...</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>The Swedish buyout firm has sold its remaining...</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text  label\n",
       "0  The GeoSolutions technology will leverage Bene...      0\n",
       "1  $ESI on lows, down $1.50 to $2.50 BK a real po...      1\n",
       "2  For the last quarter of 2010 , Componenta 's n...      0\n",
       "3  According to the Finnish-Russian Chamber of Co...      2\n",
       "4  The Swedish buyout firm has sold its remaining...      2"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = data[data['text_length'] < 61]\n",
    "\n",
    "data = data.reset_index(drop=True)\n",
    "\n",
    "data = data.drop(columns=['text_length'])\n",
    "\n",
    "data['label'] = data['label'].replace(label2id)\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert Pandas Dataframe to Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = Dataset.from_pandas(data)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert Pandas DataFrame to Dataset & Split into Train/Eval Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Dataset Shape: (8536, 2)\n",
      "Evaluation Dataset Shape: (2135, 2)\n",
      "Training Dataset Features: {'text': Value(dtype='string', id=None), 'label': Value(dtype='int64', id=None)}\n"
     ]
    }
   ],
   "source": [
    "# Split into train/eval Subsets\n",
    "train_eval = dataset.train_test_split(test_size=0.20)\n",
    "\n",
    "# Combine train/eval into datasetdict\n",
    "ds = DatasetDict({\n",
    "    'train' : train_eval['train'],\n",
    "    'eval' : train_eval['test']\n",
    "})\n",
    "\n",
    "print(f\"Training Dataset Shape: {ds['train'].shape}\")\n",
    "print(f\"Evaluation Dataset Shape: {ds['eval'].shape}\")\n",
    "print(f\"Training Dataset Features: {ds['train'].features}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Find Largest Tokenized Input Value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The longest article length is 106 tokens.\n"
     ]
    }
   ],
   "source": [
    "MODEL_CKPT = \"google/fnet-base\"\n",
    "\n",
    "tokenizer = AutoTokenizer.from_pretrained(MODEL_CKPT, \n",
    "                                             truncation=True, \n",
    "                                             padding=True,\n",
    "                                             batched=True, \n",
    "                                             batch_size=None)\n",
    "\n",
    "text_len = [len(tokenizer.encode(a)) for a in data[\"text\"]]\n",
    "\n",
    "MAX_LENGTH = max(text_len)\n",
    "\n",
    "print(f\"The longest article length is {MAX_LENGTH} tokens.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Basic Constants/Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The maximum input length is 109.\n"
     ]
    }
   ],
   "source": [
    "MODEL_CKPT = \"google/fnet-base\"\n",
    "MODEL_NAME = MODEL_CKPT.split(\"/\")[-1] + \"-Financial_Sentiment_Analysis\"\n",
    "\n",
    "DEVICE = torch.device(\"cpu\")\n",
    "MAX_INPUT_LENGTH = MAX_LENGTH + 3\n",
    "\n",
    "NUM_OF_EPOCHS = 5\n",
    "BATCH_SIZE = 64\n",
    "\n",
    "LR = 2e-5\n",
    "STRATEGY = \"epoch\"\n",
    "\n",
    "WEIGHT_DECAY = 0.01\n",
    "REPORTS_TO = \"tensorboard\"\n",
    "\n",
    "set_seed(42)\n",
    "\n",
    "print(f\"The maximum input length is {MAX_INPUT_LENGTH}.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Tokenize Entire Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bcb07138d0574d57b9b4c31c94347dd3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/8536 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4bd0bf3919f042cd8c8a4c5ddd869dfb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/2135 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Dataset Shape: (8536, 4)\n",
      "Evaluation Dataset Shape: (2135, 4)\n"
     ]
    }
   ],
   "source": [
    "# Define Tokenizer\n",
    "tokenizer = AutoTokenizer.from_pretrained(MODEL_CKPT)\n",
    "\n",
    "def tokenize(batch):\n",
    "    return tokenizer(batch['text'], \n",
    "                     truncation=True, \n",
    "                     padding=True, \n",
    "                     max_length=MAX_INPUT_LENGTH, \n",
    "                     return_tensors=\"np\")\n",
    "\n",
    "# Map 'Tokenize' Method to Entire Dataset\n",
    "encoded_ds = ds.map(tokenize, batched=True, batch_size=None)\n",
    "\n",
    "print('Training Dataset Shape:', encoded_ds['train'].shape)\n",
    "\n",
    "print('Evaluation Dataset Shape:', encoded_ds['eval'].shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Check Result of Tokenizing Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'text': Value(dtype='string', id=None), 'label': Value(dtype='int64', id=None), 'input_ids': Sequence(feature=Value(dtype='int32', id=None), length=-1, id=None), 'token_type_ids': Sequence(feature=Value(dtype='int8', id=None), length=-1, id=None)}\n",
      "Dataset({\n",
      "    features: ['text', 'label', 'input_ids', 'token_type_ids'],\n",
      "    num_rows: 8536\n",
      "})\n"
     ]
    }
   ],
   "source": [
    "print(encoded_ds['train'].features)\n",
    "print(encoded_ds['train'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Instantiate Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of the model checkpoint at google/fnet-base were not used when initializing FNetForSequenceClassification: ['cls.predictions.bias', 'cls.seq_relationship.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.decoder.weight', 'cls.predictions.decoder.bias', 'cls.predictions.transform.dense.weight', 'cls.seq_relationship.weight']\n",
      "- This IS expected if you are initializing FNetForSequenceClassification 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 FNetForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "Some weights of FNetForSequenceClassification were not initialized from the model checkpoint at google/fnet-base 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"
     ]
    },
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = AutoModelForSequenceClassification.from_pretrained(MODEL_CKPT, \n",
    "                                                           num_labels=NUM_OF_LABELS\n",
    "                                                           ).to(DEVICE)\n",
    "\n",
    "model.config.num_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Define Function to Compute Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_metrics(p):\n",
    "    accuracy_metric = evaluate.load(\"accuracy\")\n",
    "    \n",
    "    accuracy = accuracy_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                             axis=1), \n",
    "                                       references=p.label_ids)['accuracy']\n",
    "    \n",
    "    ### ------------------- F1 scores -------------------\n",
    "    \n",
    "    f1_score_metric = evaluate.load(\"f1\")\n",
    "    \n",
    "    weighted_f1_score = f1_score_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                      axis=1), \n",
    "                                                references=p.label_ids, \n",
    "                                                average='weighted')[\"f1\"]\n",
    "    \n",
    "    micro_f1_score = f1_score_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                   axis=1), \n",
    "                                             references=p.label_ids, \n",
    "                                             average='micro')['f1']\n",
    "    \n",
    "    macro_f1_score = f1_score_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                   axis=1), \n",
    "                                             references=p.label_ids, \n",
    "                                             average='macro')[\"f1\"]\n",
    "    \n",
    "    ### ------------------- recall -------------------\n",
    "    \n",
    "    recall_metric = evaluate.load(\"recall\")\n",
    "    \n",
    "    weighted_recall = recall_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                  axis=1), \n",
    "                                            references=p.label_ids, \n",
    "                                            average='weighted')[\"recall\"]\n",
    "    \n",
    "    micro_recall = recall_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                               axis=1), \n",
    "                                         references=p.label_ids, \n",
    "                                         average='micro')[\"recall\"]\n",
    "    \n",
    "    macro_recall = recall_metric.compute(predictions=np.argmax(p.predictions,\n",
    "                                                               axis=1), \n",
    "                                         references=p.label_ids, \n",
    "                                         average='macro')[\"recall\"]\n",
    "    \n",
    "    ### ------------------- precision -------------------\n",
    "    \n",
    "    precision_metric = evaluate.load(\"precision\")\n",
    "    \n",
    "    weighted_precision = precision_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                        axis=1), \n",
    "                                                  references=p.label_ids, \n",
    "                                                  average='weighted')[\"precision\"]\n",
    "    \n",
    "    micro_precision = precision_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                     axis=1), \n",
    "                                               references=p.label_ids, \n",
    "                                               average='micro')[\"precision\"]\n",
    "    \n",
    "    macro_precision = precision_metric.compute(predictions=np.argmax(p.predictions, \n",
    "                                                                     axis=1), \n",
    "                                               references=p.label_ids, \n",
    "                                               average='macro')[\"precision\"]\n",
    "    \n",
    "    # Return dictionary that includes all metrics (& their appropriate label)\n",
    "    return {\"accuracy\" : accuracy, \n",
    "            \"Weighted F1\" : weighted_f1_score,\n",
    "            \"Micro F1\" : micro_f1_score,\n",
    "            \"Macro F1\" : macro_f1_score,\n",
    "            \"Weighted Recall\" : weighted_recall,\n",
    "            \"Micro Recall\" : micro_recall,\n",
    "            \"Macro Recall\" : macro_recall,\n",
    "            \"Weighted Precision\" : weighted_precision,\n",
    "            \"Micro Precision\" : micro_precision,\n",
    "            \"Macro Precision\" : macro_precision\n",
    "            }"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Instantiate Data Collator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_collator = DataCollatorWithPadding(tokenizer=tokenizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Instantiate TrainingArguments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_args = TrainingArguments(\n",
    "    output_dir=MODEL_NAME,\n",
    "    num_train_epochs=NUM_OF_EPOCHS,\n",
    "    learning_rate=LR,\n",
    "    per_device_train_batch_size=BATCH_SIZE, \n",
    "    per_device_eval_batch_size=BATCH_SIZE, \n",
    "    weight_decay=WEIGHT_DECAY,\n",
    "    log_level=\"error\",\n",
    "    logging_strategy=\"steps\", \n",
    "    logging_steps=50,\n",
    "    evaluation_strategy=STRATEGY,\n",
    "    save_strategy=STRATEGY,\n",
    "    load_best_model_at_end=True,\n",
    "    metric_for_best_model=\"eval_Weighted F1\",\n",
    "    greater_is_better=True,\n",
    "    logging_first_step=True,\n",
    "    group_by_length=True,\n",
    "    disable_tqdm=False,\n",
    "    hub_private_repo=True,\n",
    "    push_to_hub=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Subclass Trainer to Correctly Compute Loss Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CustomTrainer(Trainer):\n",
    "    def compute_loss(self, model, inputs, return_outputs=False):\n",
    "        labels = inputs.get(\"labels\")\n",
    "        outputs = model(**inputs)\n",
    "        logits = outputs.get(\"logits\")\n",
    "        loss_fct = nn.CrossEntropyLoss(weight=torch.tensor([1.0, 2.0, 3.0]))\n",
    "        loss = loss_fct(logits.view(-1, self.model.config.num_labels), labels.view(-1))\n",
    "        return (loss, outputs) if return_outputs else loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Instantiate Trainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Cloning https://huggingface.co/DunnBC22/fnet-base-Financial_Sentiment_Analysis into local empty directory.\n"
     ]
    }
   ],
   "source": [
    "trainer = CustomTrainer(model=model, \n",
    "                  args=training_args, \n",
    "                  compute_metrics=compute_metrics,\n",
    "                  train_dataset=encoded_ds['train'], \n",
    "                  eval_dataset=encoded_ds['eval'], \n",
    "                  data_collator=data_collator,\n",
    "                  tokenizer=tokenizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/briandunn/Documents/deep_learning/dl/lib/python3.9/site-packages/transformers/optimization.py:391: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "13cd0cb614574aa7b610339ac0ea1239",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/670 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'loss': 1.3814, 'learning_rate': 1.9970149253731346e-05, 'epoch': 0.01}\n",
      "{'loss': 0.7726, 'learning_rate': 1.8507462686567165e-05, 'epoch': 0.37}\n",
      "{'loss': 0.6116, 'learning_rate': 1.701492537313433e-05, 'epoch': 0.75}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "56803b5523fa4eec9a7dbbb3040ecac5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.5126523375511169, 'eval_accuracy': 0.6304449648711944, 'eval_Weighted F1': 0.5606037561727444, 'eval_Micro F1': 0.6304449648711944, 'eval_Macro F1': 0.47045274263178705, 'eval_Weighted Recall': 0.6304449648711944, 'eval_Micro Recall': 0.6304449648711944, 'eval_Macro Recall': 0.5272363259733699, 'eval_Weighted Precision': 0.6722155428388943, 'eval_Micro Precision': 0.6304449648711944, 'eval_Macro Precision': 0.6102860649322367, 'eval_runtime': 171.7075, 'eval_samples_per_second': 12.434, 'eval_steps_per_second': 0.198, 'epoch': 1.0}\n",
      "{'loss': 0.5322, 'learning_rate': 1.5522388059701494e-05, 'epoch': 1.12}\n",
      "{'loss': 0.4776, 'learning_rate': 1.4029850746268658e-05, 'epoch': 1.49}\n",
      "{'loss': 0.4497, 'learning_rate': 1.2537313432835823e-05, 'epoch': 1.87}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "18994b072ee644e18b2f6be3d495f4b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.3885279595851898, 'eval_accuracy': 0.7578454332552693, 'eval_Weighted F1': 0.7490218274929519, 'eval_Micro F1': 0.7578454332552693, 'eval_Macro F1': 0.6783442448476253, 'eval_Weighted Recall': 0.7578454332552693, 'eval_Micro Recall': 0.7578454332552693, 'eval_Macro Recall': 0.6636389045999934, 'eval_Weighted Precision': 0.7676954272654415, 'eval_Micro Precision': 0.7578454332552693, 'eval_Macro Precision': 0.7196224211736156, 'eval_runtime': 173.0181, 'eval_samples_per_second': 12.34, 'eval_steps_per_second': 0.197, 'epoch': 2.0}\n",
      "{'loss': 0.373, 'learning_rate': 1.1044776119402986e-05, 'epoch': 2.24}\n",
      "{'loss': 0.3497, 'learning_rate': 9.552238805970149e-06, 'epoch': 2.61}\n",
      "{'loss': 0.3319, 'learning_rate': 8.059701492537314e-06, 'epoch': 2.99}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9ea3753e57c848ab95414abc58341363",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.3546285033226013, 'eval_accuracy': 0.7798594847775175, 'eval_Weighted F1': 0.7783607727205389, 'eval_Micro F1': 0.7798594847775175, 'eval_Macro F1': 0.7185041441931629, 'eval_Weighted Recall': 0.7798594847775175, 'eval_Micro Recall': 0.7798594847775175, 'eval_Macro Recall': 0.7166720710978347, 'eval_Weighted Precision': 0.7978782009461614, 'eval_Micro Precision': 0.7798594847775175, 'eval_Macro Precision': 0.7439743475218487, 'eval_runtime': 261.5408, 'eval_samples_per_second': 8.163, 'eval_steps_per_second': 0.13, 'epoch': 3.0}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Adding files tracked by Git LFS: ['.DS_Store']. This may take a bit of time if the files are large.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'loss': 0.2968, 'learning_rate': 6.567164179104478e-06, 'epoch': 3.36}\n",
      "{'loss': 0.2953, 'learning_rate': 5.074626865671642e-06, 'epoch': 3.73}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "920f86eed8a94be8ba54b114b6aa6644",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.3312208950519562, 'eval_accuracy': 0.8117096018735362, 'eval_Weighted F1': 0.8104544720135948, 'eval_Micro F1': 0.8117096018735362, 'eval_Macro F1': 0.7434939109401449, 'eval_Weighted Recall': 0.8117096018735362, 'eval_Micro Recall': 0.8117096018735362, 'eval_Macro Recall': 0.7356267558475386, 'eval_Weighted Precision': 0.8111226640479599, 'eval_Micro Precision': 0.8117096018735362, 'eval_Macro Precision': 0.7531817414593812, 'eval_runtime': 258.0356, 'eval_samples_per_second': 8.274, 'eval_steps_per_second': 0.132, 'epoch': 4.0}\n",
      "{'loss': 0.2578, 'learning_rate': 3.582089552238806e-06, 'epoch': 4.1}\n",
      "{'loss': 0.2453, 'learning_rate': 2.08955223880597e-06, 'epoch': 4.48}\n",
      "{'loss': 0.2446, 'learning_rate': 5.970149253731343e-07, 'epoch': 4.85}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a03923de3ab6426fadccd007f6849357",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.3280685544013977, 'eval_accuracy': 0.8117096018735362, 'eval_Weighted F1': 0.8110348372840417, 'eval_Micro F1': 0.8117096018735362, 'eval_Macro F1': 0.7471800669345177, 'eval_Weighted Recall': 0.8117096018735362, 'eval_Micro Recall': 0.8117096018735362, 'eval_Macro Recall': 0.7393684563239153, 'eval_Weighted Precision': 0.81435338901484, 'eval_Micro Precision': 0.8117096018735362, 'eval_Macro Precision': 0.7588001923118343, 'eval_runtime': 260.0613, 'eval_samples_per_second': 8.21, 'eval_steps_per_second': 0.131, 'epoch': 5.0}\n",
      "{'train_runtime': 15679.5405, 'train_samples_per_second': 2.722, 'train_steps_per_second': 0.043, 'train_loss': 0.3991495038146403, 'epoch': 5.0}\n"
     ]
    }
   ],
   "source": [
    "training_results = trainer.train()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Push Fine-Tuned Model to HuggingFace Hub (My Profile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Several commits (2) will be pushed upstream.\n",
      "The progress bars may be unreliable.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cae31f928c4f4bdcb401d0dcb696ae5e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Upload file pytorch_model.bin:   0%|          | 1.00/316M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0fa69f3efa584e139e9f2938b4f98641",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Upload file runs/May05_21-47-27_Brians-Mac-mini/events.out.tfevents.1683341253.Brians-Mac-mini.15242.0:   0%| …"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "To https://huggingface.co/DunnBC22/fnet-base-Financial_Sentiment_Analysis\n",
      "   bcb782d..647c2c2  main -> main\n",
      "\n",
      "To https://huggingface.co/DunnBC22/fnet-base-Financial_Sentiment_Analysis\n",
      "   647c2c2..b78a597  main -> main\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "***** train metrics *****\n",
      "  epoch                    =        5.0\n",
      "  train_loss               =     0.3991\n",
      "  train_runtime            = 4:21:19.54\n",
      "  train_samples_per_second =      2.722\n",
      "  train_steps_per_second   =      0.043\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "To https://huggingface.co/DunnBC22/fnet-base-Financial_Sentiment_Analysis\n",
      "   b78a597..1719d6b  main -> main\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'https://huggingface.co/DunnBC22/fnet-base-Financial_Sentiment_Analysis/commit/1719d6b8d6d2c3189d0161a438b89cde886d670f'"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainer.save_model()\n",
    "trainer.log_metrics(\"train\", training_results.metrics)\n",
    "trainer.save_metrics(\"train\", training_results.metrics)\n",
    "trainer.save_state()\n",
    "\n",
    "trainer.push_to_hub(commit_message=\"All Dunn!!!\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluate Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "67f832aac45d4589a2eb657fbf4231a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/34 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'eval_loss': 0.3280685544013977, 'eval_accuracy': 0.8117096018735362, 'eval_Weighted F1': 0.8110348372840417, 'eval_Micro F1': 0.8117096018735362, 'eval_Macro F1': 0.7471800669345177, 'eval_Weighted Recall': 0.8117096018735362, 'eval_Micro Recall': 0.8117096018735362, 'eval_Macro Recall': 0.7393684563239153, 'eval_Weighted Precision': 0.81435338901484, 'eval_Micro Precision': 0.8117096018735362, 'eval_Macro Precision': 0.7588001923118343, 'eval_runtime': 258.8446, 'eval_samples_per_second': 8.248, 'eval_steps_per_second': 0.131, 'epoch': 5.0}\n"
     ]
    }
   ],
   "source": [
    "metrics = trainer.evaluate()\n",
    "\n",
    "print(metrics)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes & Other Takeaways From This Project\n",
    "****\n",
    "- \n",
    "\n",
    "****"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Citation(s)\n",
    "\n",
    "- Model Checkpoint\n",
    "    > "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.12 ('nlp_ch4')",
   "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"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "a52fe47989fdc78fafbb981021cec52a6b82df6453830b9ffbd04250493e6cab"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}