{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "uF4xtbj5CRl6" }, "source": [ "# Welcome to the Edith Project!\n", "This project focuses on developing a context-based chatbot where users provide a context and can ask questions about it. The chatbot is trained using the SQuAD1.1 dataset, which provides a rich set of question-answering examples. In this notebook, we fine-tune a BERT model to adapt it to our specific task. Throughout the notebook, we will provide clear explanations for each step of the code, ensuring that readers can easily follow along and understand how the model is being prepared, trained, and evaluated. By the end, you’ll see how everything comes together to build a powerful question-answering system.\n", "\n", "## Sections:\n", "### 1. Data Preparation\n", "In this section, we will preprocess the SQuAD1.1 dataset, convert it into the right format, and tokenize the input text using BERT's tokenizer. The goal is to prepare our data efficiently for model training.\n", "\n", "### 2. Model Selection\n", "We will select the BERT model as the backbone of our chatbot and explain why this pre-trained transformer model is suitable for question-answering tasks. We’ll also load the pre-trained model and tokenizer to kick-start the fine-tuning process.\n", "\n", "### 3. Fine-Tuning and Training\n", "Here, we’ll describe how we fine-tune the BERT model for our question-answering task, covering details such as learning rates, batch sizes, and optimization steps. We will also monitor key metrics like loss and F1 score during training to gauge performance.\n", "\n", "### 4. Evaluation and Inference\n", "After training, we will evaluate the model using various metrics like Exact Match (EM) and F1 score. We’ll also demonstrate how the chatbot handles inference, where the user provides a context, and the chatbot returns the most relevant answer.\n", "\n", "### 5. Conclusion\n", "In this final section, we’ll summarize the outcomes of the project, highlighting key performance metrics and possible improvements for future iterations of the Edith Project.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19", "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5", "execution": { "iopub.execute_input": "2024-10-20T05:15:34.102347Z", "iopub.status.busy": "2024-10-20T05:15:34.101700Z", "iopub.status.idle": "2024-10-20T05:15:46.300086Z", "shell.execute_reply": "2024-10-20T05:15:46.299068Z", "shell.execute_reply.started": "2024-10-20T05:15:34.102305Z" }, "id": "IFjuS7x9CRl8" }, "outputs": [], "source": [ "!pip install -q torch transformers datasets tqdm scikit-learn rouge-score nltk datasets\n", "\n", "import re\n", "import pandas as pd\n", "import numpy as np\n", "import torch\n", "from torch.utils.data import Dataset, DataLoader\n", "from transformers import BertTokenizerFast, BertForQuestionAnswering, get_scheduler\n", "from tqdm import tqdm\n", "import os\n", "from sklearn.metrics import f1_score, precision_score, recall_score\n", "from rouge_score import rouge_scorer\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from IPython.display import display, HTML\n", "from datasets import load_dataset\n", "import seaborn as sns\n", "import matplotlib.pyplot as plt\n", "import warnings\n", "from scipy.ndimage import gaussian_filter1d\n", "\n", "warnings.filterwarnings(\"ignore\")\n" ] }, { "cell_type": "markdown", "metadata": { "id": "_lQkGREWCRl9" }, "source": [ "### 1. Data Preparation\n", "\n", "In this section, we will load our data. The dataset we are using for this project is the well-known **SQuAD1.1** (Stanford Question Answering Dataset), developed by Stanford University. SQuAD1.1 has a wide range of applications in natural language understanding and question-answering tasks. This dataset is readily available on Hugging Face, and we can load it directly using the `datasets` library.\n", "\n", "To avoid memory-related issues, we will only utilize the training dataset and split it into 95% for training and 5% for validation. The main reason for this split is that SQuAD1.1 contains approximately 87K samples, so even 5% (around 4.3K) provides a substantial number of examples for evaluation while ensuring the majority of data is used for training. This split ensures that our model has enough data to learn effectively while still being able to test its performance on a meaningful validation set.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "referenced_widgets": [ "8d0b370df64747b488f187e906b24f97", "66da7373074b464d99d90a643d56761c", "e8e4e46cc960422880747857854c489c", "f443fe30cc2b49d482045a3c65d2d468", "e8e6e53cf4974ff39ec4bdc9c6b26533" ] }, "execution": { "iopub.execute_input": "2024-10-20T05:16:33.843115Z", "iopub.status.busy": "2024-10-20T05:16:33.842244Z", "iopub.status.idle": "2024-10-20T05:16:38.885273Z", "shell.execute_reply": "2024-10-20T05:16:38.884335Z", "shell.execute_reply.started": "2024-10-20T05:16:33.843071Z" }, "id": "bz_ahmL8CRl9", "outputId": "168c6f05-61ea-484f-b096-0cdfe40eb717" }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "8d0b370df64747b488f187e906b24f97", "version_major": 2, "version_minor": 0 }, "text/plain": [ "README.md: 0%| | 0.00/7.62k [00:00\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
ContextQuestionAnswerAnswer Start IndexAnswer End Index
0Starting in the 1890s and stretching in some p...Which industries did European settlers in Alas...fishing and logging496515
1Details of school casualties had been under no...How many students were disabled in Xinhua?546450453
2Different religious traditions assign differin...What are some religious traditions that are fo...expansive powers and abilities, psychological ...120233
3The First Great Awakening was an evangelical a...What movement made a permanent mark on Protest...The First Great Awakening025
4Bacteria can be grown in the laboratory on nut...What are poultry eggs used for aside from cons...Many vaccines to infectious diseases can be gr...120197
\n", "" ], "text/plain": [ " Context \\\n", "0 Starting in the 1890s and stretching in some p... \n", "1 Details of school casualties had been under no... \n", "2 Different religious traditions assign differin... \n", "3 The First Great Awakening was an evangelical a... \n", "4 Bacteria can be grown in the laboratory on nut... \n", "\n", " Question \\\n", "0 Which industries did European settlers in Alas... \n", "1 How many students were disabled in Xinhua? \n", "2 What are some religious traditions that are fo... \n", "3 What movement made a permanent mark on Protest... \n", "4 What are poultry eggs used for aside from cons... \n", "\n", " Answer Answer Start Index \\\n", "0 fishing and logging 496 \n", "1 546 450 \n", "2 expansive powers and abilities, psychological ... 120 \n", "3 The First Great Awakening 0 \n", "4 Many vaccines to infectious diseases can be gr... 120 \n", "\n", " Answer End Index \n", "0 515 \n", "1 453 \n", "2 233 \n", "3 25 \n", "4 197 " ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Create an empty list to hold the rows\n", "train_rows = []\n", "\n", "# Iterate over the training data and collect relevant fields\n", "for data in train_data:\n", " # For each answer in the list, create a new row\n", " for answer, start_index in zip(data['answers']['text'], data['answers']['answer_start']):\n", " # Calculate the end index of the answer\n", " end_index = start_index + len(answer) if start_index is not None else 0\n", "\n", " # Append a dictionary for each entry\n", " train_rows.append({\n", " 'ID': data['id'],\n", " 'Title': data['title'],\n", " 'Context': data['context'],\n", " 'Question': data['question'],\n", " 'Answer': answer,\n", " 'Answer Start Index': start_index if start_index is not None else 0,\n", " 'Answer End Index': end_index if end_index is not None else 0\n", " })\n", "\n", "# Convert the list of dictionaries into a DataFrame\n", "train_df = pd.DataFrame(train_rows)\n", "\n", "# Replace any missing values (NaN) in \"Answer Start Index\" or \"Answer End Index\" with 0\n", "train_df['Answer Start Index'] = train_df['Answer Start Index'].fillna(0).astype(int)\n", "train_df['Answer End Index'] = train_df['Answer End Index'].fillna(0).astype(int)\n", "\n", "# Specify the columns to include in the DataFrame\n", "train_df = train_df[['Context', 'Question', 'Answer', 'Answer Start Index', 'Answer End Index']]\n", "\n", "# Display the first few rows\n", "train_df.head()\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:29:05.162471Z", "iopub.status.busy": "2024-10-20T05:29:05.161536Z", "iopub.status.idle": "2024-10-20T05:29:05.768166Z", "shell.execute_reply": "2024-10-20T05:29:05.767242Z", "shell.execute_reply.started": "2024-10-20T05:29:05.162431Z" }, "id": "absM2smvCRl-", "outputId": "2ec3e4ad-0177-48ab-f986-bffc3e508d28" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
ContextQuestionAnswerAnswer Start IndexAnswer End Index
0The Pew Forum on Religion & Public Life ranks ...What percentage of Egyptians polled support de...84%468471
1The Ann Arbor Hands-On Museum is located in a ...Ann Arbor ranks 1st among what goods sold?books402407
2One important aspect of the rule-of-law initia...In developing countries, who makes most of the...the executive612625
3In December 1547, Francis was in Malacca (Mala...Who impressed Xavier by taking notes in church?Anjiro160166
4Groups are also applied in many other mathemat...What represents elements of the fundamental gr...loops489494
\n", "
" ], "text/plain": [ " Context \\\n", "0 The Pew Forum on Religion & Public Life ranks ... \n", "1 The Ann Arbor Hands-On Museum is located in a ... \n", "2 One important aspect of the rule-of-law initia... \n", "3 In December 1547, Francis was in Malacca (Mala... \n", "4 Groups are also applied in many other mathemat... \n", "\n", " Question Answer \\\n", "0 What percentage of Egyptians polled support de... 84% \n", "1 Ann Arbor ranks 1st among what goods sold? books \n", "2 In developing countries, who makes most of the... the executive \n", "3 Who impressed Xavier by taking notes in church? Anjiro \n", "4 What represents elements of the fundamental gr... loops \n", "\n", " Answer Start Index Answer End Index \n", "0 468 471 \n", "1 402 407 \n", "2 612 625 \n", "3 160 166 \n", "4 489 494 " ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Create an empty list to hold the rows\n", "val_rows = []\n", "\n", "# Iterate over the training data and collect relevant fields\n", "for data in val_data:\n", " # For each answer in the list, create a new row\n", " for answer, start_index in zip(data['answers']['text'], data['answers']['answer_start']):\n", " # Calculate the end index of the answer\n", " end_index = start_index + len(answer) if start_index is not None else 0\n", "\n", " # Append a dictionary for each entry\n", " val_rows.append({\n", " 'ID': data['id'],\n", " 'Title': data['title'],\n", " 'Context': data['context'],\n", " 'Question': data['question'],\n", " 'Answer': answer,\n", " 'Answer Start Index': start_index if start_index is not None else 0,\n", " 'Answer End Index': end_index if end_index is not None else 0\n", " })\n", "\n", "# Convert the list of dictionaries into a DataFrame\n", "val_df = pd.DataFrame(val_rows)\n", "\n", "# Replace any missing values (NaN) in \"Answer Start Index\" or \"Answer End Index\" with 0\n", "val_df['Answer Start Index'] = val_df['Answer Start Index'].fillna(0).astype(int)\n", "val_df['Answer End Index'] = val_df['Answer End Index'].fillna(0).astype(int)\n", "\n", "# Specify the columns to include in the DataFrame\n", "val_df = val_df[['Context', 'Question', 'Answer', 'Answer Start Index', 'Answer End Index']]\n", "\n", "# Display the first few rows\n", "val_df.head()\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:30:08.174343Z", "iopub.status.busy": "2024-10-20T05:30:08.173933Z", "iopub.status.idle": "2024-10-20T05:30:11.267282Z", "shell.execute_reply": "2024-10-20T05:30:11.266457Z", "shell.execute_reply.started": "2024-10-20T05:30:08.174305Z" }, "id": "H-5ifapQCRl-" }, "outputs": [], "source": [ "train_df.to_csv(\"squad_train.csv\", index=False)\n", "val_df.to_csv(\"squad_val.csv\", index=False)\n" ] }, { "cell_type": "markdown", "metadata": { "id": "hG5pwTKsCRl-" }, "source": [ "1. **Loading the SQuAD1.1 dataset**: The code begins by loading the SQuAD1.1 dataset using the `load_dataset` function from the `datasets` library. The training set (`train_data`) is then split into training and validation sets using an 95:5 ratio. This ensures that a portion of the original training data is reserved for validation purposes during model training.\n", "\n", "2. **Initializing lists for storing data**: An empty list, `train_rows`, is created to store the processed rows of data from the training set. This will later be converted into a DataFrame for easier manipulation and saving.\n", "\n", "3. **Processing training data**: The code iterates over each entry in the `train_data` split. For each entry, it extracts the ID, title, context, question, and answers from the dataset. For each answer, the corresponding start and end indices are calculated and stored in a dictionary.\n", "\n", "4. **Converting training data to DataFrame**: Once all rows of the training data have been processed, they are converted into a pandas DataFrame. This allows for easier data manipulation, including replacing missing values and ensuring that the start and end indices are integers.\n", "\n", "5. **Processing validation data**: The validation data undergoes the same process as the training data. The relevant fields (ID, title, context, question, and answers) are extracted, the start and end indices are calculated, and the data is stored in a list of dictionaries.\n", "\n", "6. **Converting validation data to DataFrame**: Similar to the training data, the processed validation data is converted into a pandas DataFrame. Any missing values in the start and end indices are replaced with zeroes, and the columns are organized for consistency.\n", "\n", "7. **Saving data to CSV**: Finally, the training and validation DataFrames are saved as CSV files (`squad_train.csv` and `squad_val.csv`). This makes it easy to load the preprocessed data for further analysis or model training.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:18.689006Z", "iopub.status.busy": "2024-10-20T05:42:18.688153Z", "iopub.status.idle": "2024-10-20T05:42:18.870198Z", "shell.execute_reply": "2024-10-20T05:42:18.869197Z", "shell.execute_reply.started": "2024-10-20T05:42:18.688967Z" }, "id": "KXHxa1-_CRl-" }, "outputs": [], "source": [ "# Create a custom dataset class for our data\n", "class SQuADataset(Dataset):\n", " def __init__(self, df, tokenizer, max_len):\n", " self.df = df\n", " self.tokenizer = tokenizer\n", " self.max_len = max_len\n", "\n", " def __len__(self):\n", " return self.df.shape[0]\n", "\n", " def __getitem__(self, idx):\n", " context = self.df.iloc[idx]['Context']\n", " question = self.df.iloc[idx]['Question']\n", " answer = self.df.iloc[idx]['Answer']\n", " answer_start = self.df.iloc[idx]['Answer Start Index']\n", "\n", " encoding = self.tokenizer.encode_plus(\n", " context,\n", " question,\n", " max_length=self.max_len,\n", " truncation=True,\n", " padding='max_length',\n", " return_offsets_mapping=True,\n", " return_token_type_ids=True,\n", " return_attention_mask=True,\n", " return_tensors='pt'\n", " )\n", "\n", " offset_mapping = encoding['offset_mapping'].squeeze().tolist()\n", "\n", " # Initialize start and end positions of the answer\n", " start_position = 0\n", " end_position = 0\n", "\n", " # Find the token positions corresponding to the answer if it's not unanswerable\n", " if answer_start != -1:\n", " for idx, (start, end) in enumerate(offset_mapping):\n", " if start <= answer_start < end:\n", " start_position = idx\n", " if start < answer_start + len(answer) <= end:\n", " end_position = idx\n", " break\n", "\n", " return {\n", " 'input_ids': encoding['input_ids'].squeeze(),\n", " 'attention_mask': encoding['attention_mask'].squeeze(),\n", " 'token_type_ids': encoding['token_type_ids'].squeeze(),\n", " 'start_positions': torch.tensor(start_position),\n", " 'end_positions': torch.tensor(end_position)\n", " }\n", "\n", "\n", "# Set up the tokenizer and model\n", "tokenizer = BertTokenizerFast.from_pretrained('bert-base-uncased')\n", "\n", "# Create the dataset and data loader\n", "train_dataset = SQuADataset(train_df, tokenizer, max_len=256)\n", "val_dataset = SQuADataset(val_df, tokenizer, max_len=256)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:19.426931Z", "iopub.status.busy": "2024-10-20T05:42:19.426035Z", "iopub.status.idle": "2024-10-20T05:42:19.441008Z", "shell.execute_reply": "2024-10-20T05:42:19.439956Z", "shell.execute_reply.started": "2024-10-20T05:42:19.426887Z" }, "id": "HKSWQwK-CRl_", "outputId": "8f7d41d1-41dc-4683-aa56-fd43a208f98a" }, "outputs": [ { "data": { "text/plain": [ "{'input_ids': tensor([ 101, 3225, 1999, 1996, 13678, 1998, 10917, 1999, 2070, 3182,\n", " 2000, 1996, 2220, 28088, 1010, 2751, 18545, 1999, 7397, 1998,\n", " 1996, 3518, 19898, 3700, 2716, 5190, 1997, 11257, 1998, 7322,\n", " 2000, 7397, 1012, 7397, 2001, 3985, 5100, 2004, 2019, 4114,\n", " 3700, 1999, 4878, 1012, 7397, 1005, 1055, 3007, 1010, 2029,\n", " 2018, 2042, 1999, 4133, 2912, 2127, 5518, 1010, 2001, 2333,\n", " 2167, 2000, 2238, 4887, 1012, 2810, 1997, 1996, 7397, 3099,\n", " 1005, 1055, 7330, 2211, 2008, 2168, 2095, 1012, 2647, 7489,\n", " 2013, 5120, 1998, 4701, 2036, 3876, 1999, 4643, 7397, 1010,\n", " 2073, 2027, 3133, 1996, 5645, 1998, 15899, 6088, 1012, 102,\n", " 2029, 6088, 2106, 2647, 7322, 1999, 7397, 4088, 1029, 102,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0]),\n", " 'attention_mask': tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]),\n", " 'token_type_ids': tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]),\n", " 'start_positions': tensor(94),\n", " 'end_positions': tensor(96)}" ] }, "execution_count": 37, "metadata": {}, "output_type": "execute_result" } ], "source": [ "train_dataset[0]" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:20.436694Z", "iopub.status.busy": "2024-10-20T05:42:20.436305Z", "iopub.status.idle": "2024-10-20T05:42:20.441943Z", "shell.execute_reply": "2024-10-20T05:42:20.440869Z", "shell.execute_reply.started": "2024-10-20T05:42:20.436643Z" }, "id": "WAeJpg0wCRl_" }, "outputs": [], "source": [ "train_loader = DataLoader(train_dataset, batch_size=32, shuffle=True)\n", "val_loader = DataLoader(val_dataset, batch_size=32, shuffle=False)" ] }, { "cell_type": "markdown", "metadata": { "id": "HOkWDbSnCRl_" }, "source": [ "1. **Creating a Custom Dataset Class**: This code defines a custom dataset class called `SQuADataset`, which is built on top of PyTorch's `Dataset` class. This class is specifically designed to handle data from the SQuAD dataset, making it easier to load and work with during model training. It takes a DataFrame with context, questions, and answers, along with a tokenizer and a maximum sequence length as its parameters.\n", "\n", "2. **Implementing the `__len__` Method**: The `__len__` method is implemented to return the total number of samples in the DataFrame. This is important for the DataLoader to know how many items it can iterate over when creating batches.\n", "\n", "3. **Implementing the `__getitem__` Method**: In the `__getitem__` method, we retrieve the data for a specific index. This includes extracting the context, question, answer, and the starting position of the answer from the DataFrame. This method is crucial because it enables the DataLoader to access each sample during training.\n", "\n", "4. **Encoding Context and Question**: Here, the `tokenizer.encode_plus` function is used to convert the context and question into a format suitable for the BERT model. It generates input IDs, attention masks, and token type IDs, ensuring that the sequence length does not exceed the specified maximum. This encoding step is essential for preparing the text for the model.\n", "\n", "5. **Extracting Offset Mappings**: The encoded data includes offset mappings, which help us locate where in the original text the tokens correspond. This is particularly useful for identifying the exact position of the answer within the context.\n", "\n", "6. **Determining Answer Positions**: The code checks if the answer is unanswerable (indicated by `answer_start` being `-1`). If there is an answer, it uses the offset mappings to find the token positions that correspond to the answer in the context. These start and end positions will be used later for training the model.\n", "\n", "7. **Returning Encoded Inputs**: The `__getitem__` method returns a dictionary that contains the encoded input IDs, attention masks, token type IDs, and the start and end positions of the answer as tensors. This format is ready for input into the BERT model, simplifying the training process.\n", "\n", "8. **Setting Up the Tokenizer**: The code initializes a tokenizer using `BertTokenizerFast`, which is known for its speed and efficiency. This tokenizer will be used to prepare the context and questions in our dataset.\n", "\n", "9. **Creating Dataset Instances**: Two instances of the `SQuADataset` class are created: one for the training data and another for validation. The maximum sequence length is set to 256 tokens, which is a common practice to ensure a balance between memory usage and performance.\n", "\n", "10. **Setting Up Data Loaders**: Finally, the code sets up `DataLoader` instances for both the training and validation datasets. The training DataLoader shuffles the data to provide the model with a varied set of examples in each epoch, while the validation DataLoader maintains the order for evaluation. These DataLoaders make batch processing during training and evaluation efficient and straightforward.\n", "\n", "Overall, this code lays the groundwork for preparing the SQuAD dataset for training a BERT model, ensuring that the input data is properly formatted and that the answer positions are accurately identified for effective question answering.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "6MRByJNaCRl_" }, "source": [ "### 2. Model Selection\n", "\n", "In this section, we are selecting the BERT model, which stands for **Bidirectional Encoder Representations from Transformers**. BERT was developed by Google and is one of the most powerful models for natural language understanding tasks. What sets BERT apart from previous models is its **bidirectional training of transformers**—meaning it looks at the entire context (both left and right) of a word in a sentence, rather than just reading text from left to right or right to left. This allows it to gain a much deeper understanding of the context in which a word appears.\n", "\n", "BERT was pre-trained on a massive amount of text data, including Wikipedia and books, using two key tasks:\n", "1. **Masked Language Modeling (MLM)**: Where random words in a sentence are masked, and the model learns to predict them based on the surrounding context. This allows BERT to capture relationships between words.\n", "2. **Next Sentence Prediction (NSP)**: Where BERT learns to predict whether two sentences follow each other in the text, which helps the model grasp the relationships between different sentences.\n", "\n", "For our task of **Question Answering**, we use `BertForQuestionAnswering`, a version of BERT that has been fine-tuned for answering questions given a context. The model takes both the context and question as input and predicts the start and end positions of the answer within the context. This makes BERT an excellent choice for our project, as its deep contextual understanding enables it to pinpoint the answer even when the relationship between the question and context is complex.\n", "\n", "In this particular code block, the line `model = BertForQuestionAnswering.from_pretrained('bert-base-uncased')` loads a pre-trained version of BERT specifically fine-tuned for the question-answering task. The `bert-base-uncased` version uses a vocabulary where all words are lowercased, which simplifies the task of dealing with different word cases in the input text.\n", "\n", "Next, we move the model to the GPU using `model.to(device)`. This step ensures that if a GPU (CUDA) is available, the model will use it for faster computation, which is important for both training and inference. If no GPU is available, the model will default to the CPU, though it will be slower.\n", "\n", "In summary, BERT is a great fit for our task because:\n", "- Its bidirectional nature helps it understand the question and context deeply.\n", "- It has already been fine-tuned for question-answering tasks, saving us the effort of training from scratch.\n", "- It's robust and has been proven to perform well on tasks like SQuAD, which is the dataset we are using.\n", "\n", "By leveraging BERT, we can expect strong performance on our question-answering task, allowing the model to identify precise answers within a given context.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:21.891350Z", "iopub.status.busy": "2024-10-20T05:42:21.890979Z", "iopub.status.idle": "2024-10-20T05:42:22.349968Z", "shell.execute_reply": "2024-10-20T05:42:22.348904Z", "shell.execute_reply.started": "2024-10-20T05:42:21.891314Z" }, "id": "xKM-qTWdCRl_", "outputId": "b04e78e1-02fd-447a-b7c9-7e925af7763e" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "Some weights of BertForQuestionAnswering were not initialized from the model checkpoint at bert-base-uncased and are newly initialized: ['qa_outputs.bias', 'qa_outputs.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": [ "BertForQuestionAnswering(\n", " (bert): BertModel(\n", " (embeddings): BertEmbeddings(\n", " (word_embeddings): Embedding(30522, 768, padding_idx=0)\n", " (position_embeddings): Embedding(512, 768)\n", " (token_type_embeddings): Embedding(2, 768)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " (encoder): BertEncoder(\n", " (layer): ModuleList(\n", " (0-11): 12 x BertLayer(\n", " (attention): BertAttention(\n", " (self): BertSdpaSelfAttention(\n", " (query): Linear(in_features=768, out_features=768, bias=True)\n", " (key): Linear(in_features=768, out_features=768, bias=True)\n", " (value): Linear(in_features=768, out_features=768, bias=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " (output): BertSelfOutput(\n", " (dense): Linear(in_features=768, out_features=768, bias=True)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " )\n", " (intermediate): BertIntermediate(\n", " (dense): Linear(in_features=768, out_features=3072, bias=True)\n", " (intermediate_act_fn): GELUActivation()\n", " )\n", " (output): BertOutput(\n", " (dense): Linear(in_features=3072, out_features=768, bias=True)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " )\n", " )\n", " )\n", " )\n", " (qa_outputs): Linear(in_features=768, out_features=2, bias=True)\n", ")" ] }, "execution_count": 40, "metadata": {}, "output_type": "execute_result" } ], "source": [ "device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')\n", "device\n", "\n", "model = BertForQuestionAnswering.from_pretrained('bert-base-uncased')\n", "model.to(device)" ] }, { "cell_type": "markdown", "metadata": { "id": "hrN0kscZCRmA" }, "source": [ "### 3. Fine-Tuning and Training\n", "\n", "**Pre-Code Explanation**:\n", "In this section, we define the function that will fine-tune the BERT model for our specific question-answering task. Fine-tuning a pre-trained model like BERT involves adjusting the weights of the model slightly to specialize it on a new dataset. We start by setting hyperparameters such as the number of epochs, the learning rate (`3e-5`), and the optimizer. In this case, we use the `AdamW` optimizer, which is a variation of Adam designed to work well with transformers like BERT by incorporating weight decay to reduce overfitting.\n", "\n", "A **scheduler** is set up to adjust the learning rate gradually over the training process using the `get_scheduler` function, which linearly decays the learning rate as training progresses. This helps maintain model stability during training and prevents drastic fluctuations in the loss. The function accepts the training and validation datasets, performs training in epochs, and logs important statistics such as loss values to track the model's performance.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:57.738843Z", "iopub.status.busy": "2024-10-20T05:42:57.738090Z", "iopub.status.idle": "2024-10-20T05:42:57.750717Z", "shell.execute_reply": "2024-10-20T05:42:57.749771Z", "shell.execute_reply.started": "2024-10-20T05:42:57.738792Z" }, "id": "g6atLKO1CRmA" }, "outputs": [], "source": [ "# Define the training function\n", "def train_model(model, training_data, validation_data, epochs=5, learning_rate=3e-5):\n", " loss_history = [] # Store loss values\n", " optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n", "\n", " # Set up the learning rate scheduler\n", " total_steps = epochs * len(training_data)\n", " scheduler = get_scheduler(\n", " name=\"linear\", optimizer=optimizer, num_warmup_steps=0, num_training_steps=total_steps\n", " )\n", "\n", " for ep in range(epochs): # Loop through epochs\n", " model.train()\n", " epoch_train_loss = 0\n", " train_progress = tqdm(training_data, desc=\"Training\")\n", "\n", " for step_idx, step_batch in enumerate(train_progress):\n", " step_batch = {key: value.to(device) for key, value in step_batch.items()}\n", " output = model(**step_batch)\n", " loss_value = output.loss\n", " epoch_train_loss += loss_value.item()\n", " loss_history.append(loss_value.item())\n", "\n", " optimizer.zero_grad()\n", " loss_value.backward()\n", " optimizer.step()\n", " scheduler.step()\n", "\n", " # Print the average loss for the last 100 batches\n", " if (step_idx + 1) % 100 == 0:\n", " avg_last_100 = sum(loss_history[-100:]) / len(loss_history[-100:])\n", " print(f\"Avg loss for last 100 steps (step {step_idx + 1}): {avg_last_100}\")\n", "\n", " # Update progress bar with the current loss\n", " train_progress.set_postfix({\"loss\": loss_value.item()})\n", "\n", " average_train_loss = epoch_train_loss / len(training_data)\n", " print(f\"Average Training Loss for Epoch {ep + 1}: {average_train_loss}\")\n", "\n", "\n", " model.eval()\n", " total_validation_loss = 0\n", " validation_progress = tqdm(validation_data, desc=\"Validation\")\n", "\n", " for val_idx, val_batch in enumerate(validation_progress):\n", " val_batch = {key: value.to(device) for key, value in val_batch.items()}\n", " with torch.no_grad():\n", " output = model(**val_batch)\n", " val_loss = output.loss\n", " total_validation_loss += val_loss.item()\n", "\n", " average_validation_loss = total_validation_loss / len(validation_data)\n", " print(f\"Average Validation Loss for Epoch {ep + 1}: {average_validation_loss}\")\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T05:42:58.376954Z", "iopub.status.busy": "2024-10-20T05:42:58.376364Z", "iopub.status.idle": "2024-10-20T11:14:12.678036Z", "shell.execute_reply": "2024-10-20T11:14:12.676408Z", "shell.execute_reply.started": "2024-10-20T05:42:58.376902Z" }, "id": "Cvywll7_CRmA", "outputId": "9f0415ca-48a9-4d43-e2da-e632e388feea" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "Training: 4%|▍ | 99/2601 [02:28<1:02:40, 1.50s/it, loss=1.93]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 100): 2.9181237602233887\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 8%|▊ | 199/2601 [04:58<1:00:01, 1.50s/it, loss=1.58]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 200): 1.843290798664093\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 11%|█▏ | 299/2601 [07:28<57:40, 1.50s/it, loss=1.44] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 300): 1.6185305523872375\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 15%|█▌ | 399/2601 [09:58<55:06, 1.50s/it, loss=1.56] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 400): 1.5289679777622223\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 19%|█▉ | 499/2601 [12:28<52:31, 1.50s/it, loss=1.72] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 500): 1.447012750506401\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 23%|██▎ | 599/2601 [14:58<50:03, 1.50s/it, loss=1.71] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 600): 1.3697541028261184\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 27%|██▋ | 699/2601 [17:28<47:32, 1.50s/it, loss=1.34] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 700): 1.3844239920377732\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 31%|███ | 799/2601 [19:58<45:05, 1.50s/it, loss=1.19] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 800): 1.2496907639503478\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 35%|███▍ | 899/2601 [22:28<42:47, 1.51s/it, loss=1.33] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 900): 1.2965397924184798\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 38%|███▊ | 999/2601 [24:58<40:03, 1.50s/it, loss=1.28] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1000): 1.2438341867923737\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 42%|████▏ | 1099/2601 [27:28<37:34, 1.50s/it, loss=0.851]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1100): 1.2373633629083634\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 46%|████▌ | 1199/2601 [29:58<35:01, 1.50s/it, loss=1.28] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1200): 1.1354780173301697\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 50%|████▉ | 1299/2601 [32:29<32:41, 1.51s/it, loss=1.13] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1300): 1.1638590478897095\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 54%|█████▍ | 1399/2601 [34:59<30:03, 1.50s/it, loss=0.963]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1400): 1.1318764191865922\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 58%|█████▊ | 1499/2601 [37:29<27:33, 1.50s/it, loss=0.97] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1500): 1.1282511121034622\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 61%|██████▏ | 1599/2601 [39:59<25:03, 1.50s/it, loss=1.43] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1600): 1.1512188524007798\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 65%|██████▌ | 1699/2601 [42:29<22:34, 1.50s/it, loss=0.924]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1700): 1.1037893587350844\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 69%|██████▉ | 1799/2601 [45:00<20:01, 1.50s/it, loss=1.37] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1800): 1.1223931908607483\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 73%|███████▎ | 1899/2601 [47:30<17:32, 1.50s/it, loss=1.22] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1900): 1.0834963840246201\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 77%|███████▋ | 1999/2601 [50:00<15:02, 1.50s/it, loss=1.3] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2000): 1.0893496036529542\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 81%|████████ | 2099/2601 [52:30<12:33, 1.50s/it, loss=0.894]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2100): 1.0850343543291092\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 85%|████████▍ | 2199/2601 [55:00<10:04, 1.50s/it, loss=0.973]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2200): 1.098406316637993\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 88%|████████▊ | 2299/2601 [57:30<07:33, 1.50s/it, loss=0.931]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2300): 1.0876016092300416\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 92%|█████████▏| 2399/2601 [1:00:00<05:02, 1.50s/it, loss=1.54]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2400): 1.0816582387685776\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 96%|█████████▌| 2499/2601 [1:02:30<02:33, 1.50s/it, loss=1.12] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2500): 1.0321623659133912\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|█████████▉| 2599/2601 [1:05:00<00:03, 1.50s/it, loss=1.02] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2600): 1.0567743647098542\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|██████████| 2601/2601 [1:05:03<00:00, 1.50s/it, loss=1.02] \n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Training Loss for Epoch 1: 1.295620406077303\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Validation: 100%|██████████| 137/137 [01:12<00:00, 1.90it/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Validation Loss for Epoch 1: 0.9500351820113885\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 4%|▍ | 99/2601 [02:28<1:02:41, 1.50s/it, loss=0.697]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 100): 0.8142830759286881\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 8%|▊ | 199/2601 [04:58<1:00:13, 1.50s/it, loss=0.865]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 200): 0.8329307380318641\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 11%|█▏ | 299/2601 [07:29<57:33, 1.50s/it, loss=1.08] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 300): 0.7891313913464546\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 15%|█▌ | 399/2601 [09:59<55:04, 1.50s/it, loss=0.666]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 400): 0.7705841365456582\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 19%|█▉ | 499/2601 [12:29<52:39, 1.50s/it, loss=1.27] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 500): 0.8177345561981201\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 23%|██▎ | 599/2601 [14:59<50:00, 1.50s/it, loss=1.09] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 600): 0.8327104163169861\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 27%|██▋ | 699/2601 [17:29<47:32, 1.50s/it, loss=0.866]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 700): 0.810658627152443\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 31%|███ | 799/2601 [19:59<45:09, 1.50s/it, loss=0.88] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 800): 0.8449877372384071\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 35%|███▍ | 899/2601 [22:30<42:32, 1.50s/it, loss=0.661]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 900): 0.8116492775082588\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 38%|███▊ | 999/2601 [25:00<40:05, 1.50s/it, loss=1.19] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1000): 0.8410855442285537\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 42%|████▏ | 1099/2601 [27:30<37:36, 1.50s/it, loss=1.02] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1100): 0.804014807343483\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 46%|████▌ | 1199/2601 [30:00<35:11, 1.51s/it, loss=0.962]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1200): 0.8255293396115303\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 50%|████▉ | 1299/2601 [32:30<32:29, 1.50s/it, loss=0.837]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1300): 0.8005913615226745\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 54%|█████▍ | 1399/2601 [35:00<30:02, 1.50s/it, loss=1.16] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1400): 0.7952284491062165\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 58%|█████▊ | 1499/2601 [37:30<27:38, 1.51s/it, loss=0.865]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1500): 0.7911309421062469\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 61%|██████▏ | 1599/2601 [40:01<25:07, 1.50s/it, loss=0.47] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1600): 0.8090464314818382\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 65%|██████▌ | 1699/2601 [42:31<22:34, 1.50s/it, loss=0.98] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1700): 0.7942702674865723\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 69%|██████▉ | 1799/2601 [45:01<20:00, 1.50s/it, loss=0.549]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1800): 0.7940633481740952\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 73%|███████▎ | 1899/2601 [47:31<17:36, 1.50s/it, loss=0.664]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1900): 0.7622592401504517\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 77%|███████▋ | 1999/2601 [50:01<15:07, 1.51s/it, loss=0.791]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2000): 0.8292095738649369\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 81%|████████ | 2099/2601 [52:32<12:34, 1.50s/it, loss=0.874]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2100): 0.8038192576169968\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 85%|████████▍ | 2199/2601 [55:02<10:02, 1.50s/it, loss=0.98] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2200): 0.8064938718080521\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 88%|████████▊ | 2299/2601 [57:32<07:32, 1.50s/it, loss=1.25] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2300): 0.8164249670505523\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 92%|█████████▏| 2399/2601 [1:00:02<05:02, 1.50s/it, loss=1.06] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2400): 0.7864927875995636\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 96%|█████████▌| 2499/2601 [1:02:32<02:33, 1.50s/it, loss=0.497]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2500): 0.8313259115815163\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|█████████▉| 2599/2601 [1:05:02<00:02, 1.49s/it, loss=0.522]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2600): 0.7982275319099427\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|██████████| 2601/2601 [1:05:04<00:00, 1.50s/it, loss=0.727]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Training Loss for Epoch 2: 0.8081949698562945\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Validation: 100%|██████████| 137/137 [01:12<00:00, 1.90it/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Validation Loss for Epoch 2: 0.9094252549383762\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 4%|▍ | 99/2601 [02:28<1:02:36, 1.50s/it, loss=0.409]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 100): 0.5861263217031956\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 8%|▊ | 199/2601 [04:59<1:00:14, 1.50s/it, loss=0.527]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 200): 0.5470401339232922\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 11%|█▏ | 299/2601 [07:29<57:43, 1.50s/it, loss=0.637] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 300): 0.5824183216691017\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 15%|█▌ | 399/2601 [09:59<55:12, 1.50s/it, loss=0.592]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 400): 0.5499672368168831\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 19%|█▉ | 499/2601 [12:29<52:39, 1.50s/it, loss=0.695]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 500): 0.5772905349731445\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 23%|██▎ | 599/2601 [14:59<50:05, 1.50s/it, loss=1.13] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 600): 0.5968974930047989\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 27%|██▋ | 699/2601 [17:29<47:28, 1.50s/it, loss=0.443]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 700): 0.5567461925745011\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 31%|███ | 799/2601 [19:59<45:03, 1.50s/it, loss=0.395]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 800): 0.5894309616088867\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 35%|███▍ | 899/2601 [22:29<42:34, 1.50s/it, loss=0.771]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 900): 0.5873063451051712\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 38%|███▊ | 999/2601 [24:59<40:05, 1.50s/it, loss=0.476]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1000): 0.5972063279151917\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 42%|████▏ | 1099/2601 [27:29<37:30, 1.50s/it, loss=0.565]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1100): 0.6006231245398521\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 46%|████▌ | 1199/2601 [29:59<35:02, 1.50s/it, loss=0.472]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1200): 0.5915874150395394\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 50%|████▉ | 1299/2601 [32:29<32:33, 1.50s/it, loss=0.532]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1300): 0.5902719554305077\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 54%|█████▍ | 1399/2601 [35:00<30:05, 1.50s/it, loss=0.443]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1400): 0.580033713877201\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 58%|█████▊ | 1499/2601 [37:30<27:34, 1.50s/it, loss=0.808]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1500): 0.5817161786556244\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 61%|██████▏ | 1599/2601 [40:00<25:03, 1.50s/it, loss=0.733]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1600): 0.5787480530142785\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 65%|██████▌ | 1699/2601 [42:30<22:31, 1.50s/it, loss=0.44] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1700): 0.5743653793632985\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 69%|██████▉ | 1799/2601 [45:00<20:03, 1.50s/it, loss=0.373]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1800): 0.5669003206491471\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 73%|███████▎ | 1899/2601 [47:30<17:35, 1.50s/it, loss=0.441]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1900): 0.5621130636334419\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 77%|███████▋ | 1999/2601 [50:00<15:02, 1.50s/it, loss=0.396]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2000): 0.592484669983387\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 81%|████████ | 2099/2601 [52:30<12:34, 1.50s/it, loss=0.507]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2100): 0.5556239295005798\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 85%|████████▍ | 2199/2601 [55:00<10:04, 1.50s/it, loss=0.486]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2200): 0.571899283528328\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 88%|████████▊ | 2299/2601 [57:31<07:35, 1.51s/it, loss=0.279]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2300): 0.5776233053207398\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 92%|█████████▏| 2399/2601 [1:00:01<05:02, 1.50s/it, loss=0.65]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2400): 0.5737731790542603\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 96%|█████████▌| 2499/2601 [1:02:31<02:32, 1.50s/it, loss=0.893]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2500): 0.5883317263424397\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|█████████▉| 2599/2601 [1:05:01<00:02, 1.50s/it, loss=0.677]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2600): 0.575586271584034\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|██████████| 2601/2601 [1:05:03<00:00, 1.50s/it, loss=0.634]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Training Loss for Epoch 3: 0.5781794790692716\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Validation: 100%|██████████| 137/137 [01:12<00:00, 1.90it/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Validation Loss for Epoch 3: 0.999135330210637\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 4%|▍ | 99/2601 [02:28<1:02:43, 1.50s/it, loss=0.285]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 100): 0.413784771412611\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 8%|▊ | 199/2601 [04:59<1:00:03, 1.50s/it, loss=0.432]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 200): 0.43284411162137987\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 11%|█▏ | 299/2601 [07:29<57:39, 1.50s/it, loss=0.355] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 300): 0.44033734187483786\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 15%|█▌ | 399/2601 [09:59<55:03, 1.50s/it, loss=0.457] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 400): 0.4177935104072094\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 19%|█▉ | 499/2601 [12:29<52:36, 1.50s/it, loss=0.262]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 500): 0.42030997440218926\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 23%|██▎ | 599/2601 [14:59<50:10, 1.50s/it, loss=0.417]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 600): 0.43777878060936926\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 27%|██▋ | 699/2601 [17:29<47:39, 1.50s/it, loss=0.252]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 700): 0.44543042451143267\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 31%|███ | 799/2601 [19:59<45:02, 1.50s/it, loss=0.331]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 800): 0.41820422530174256\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 35%|███▍ | 899/2601 [22:29<42:16, 1.49s/it, loss=0.434]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 900): 0.4102541197836399\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 38%|███▊ | 999/2601 [24:59<40:03, 1.50s/it, loss=0.381]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1000): 0.4401281327009201\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 42%|████▏ | 1099/2601 [27:29<37:33, 1.50s/it, loss=0.375]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1100): 0.40734992161393163\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 46%|████▌ | 1199/2601 [29:59<35:01, 1.50s/it, loss=0.347]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1200): 0.4229107615351677\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 50%|████▉ | 1299/2601 [32:30<32:35, 1.50s/it, loss=0.596]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1300): 0.4366910111904144\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 54%|█████▍ | 1399/2601 [35:00<30:02, 1.50s/it, loss=0.514]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1400): 0.41040504559874535\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 58%|█████▊ | 1499/2601 [37:30<27:33, 1.50s/it, loss=0.321]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1500): 0.41208766609430314\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 61%|██████▏ | 1599/2601 [40:00<25:02, 1.50s/it, loss=0.362]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1600): 0.4288914044201374\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 65%|██████▌ | 1699/2601 [42:30<22:39, 1.51s/it, loss=0.348]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1700): 0.4072799864411354\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 69%|██████▉ | 1799/2601 [45:00<20:03, 1.50s/it, loss=0.18] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1800): 0.40538637042045594\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 73%|███████▎ | 1899/2601 [47:30<17:32, 1.50s/it, loss=0.394]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1900): 0.41915873721241953\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 77%|███████▋ | 1999/2601 [50:00<15:00, 1.50s/it, loss=0.36] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2000): 0.42285729214549067\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 81%|████████ | 2099/2601 [52:30<12:32, 1.50s/it, loss=0.45] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2100): 0.3937626303732395\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 85%|████████▍ | 2199/2601 [55:00<10:03, 1.50s/it, loss=0.365]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2200): 0.4223772123456001\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 88%|████████▊ | 2299/2601 [57:30<07:33, 1.50s/it, loss=0.447]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2300): 0.4191547580063343\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 92%|█████████▏| 2399/2601 [1:00:00<05:03, 1.50s/it, loss=0.423]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2400): 0.4223488415777683\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 96%|█████████▌| 2499/2601 [1:02:30<02:33, 1.50s/it, loss=0.597]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2500): 0.4321207369863987\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|█████████▉| 2599/2601 [1:05:00<00:02, 1.50s/it, loss=0.288]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2600): 0.4203851442039013\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|██████████| 2601/2601 [1:05:03<00:00, 1.50s/it, loss=0.632]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Training Loss for Epoch 4: 0.4216204704870769\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Validation: 100%|██████████| 137/137 [01:12<00:00, 1.90it/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Validation Loss for Epoch 4: 1.0965224147713097\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 4%|▍ | 99/2601 [02:28<1:02:26, 1.50s/it, loss=0.324] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 100): 0.31893704675137996\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 8%|▊ | 199/2601 [04:58<1:00:02, 1.50s/it, loss=0.446]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 200): 0.30915156394243243\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 11%|█▏ | 299/2601 [07:28<57:27, 1.50s/it, loss=0.232] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 300): 0.32221948117017746\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 15%|█▌ | 399/2601 [09:58<54:44, 1.49s/it, loss=0.185]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 400): 0.32024823017418386\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 19%|█▉ | 499/2601 [12:28<52:28, 1.50s/it, loss=0.563]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 500): 0.3506594298779964\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 23%|██▎ | 599/2601 [14:57<49:54, 1.50s/it, loss=0.215] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 600): 0.3250641016662121\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 27%|██▋ | 699/2601 [17:27<47:40, 1.50s/it, loss=0.338]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 700): 0.3352129091322422\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 31%|███ | 799/2601 [19:57<45:01, 1.50s/it, loss=0.209]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 800): 0.3395147521793842\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 35%|███▍ | 899/2601 [22:27<42:32, 1.50s/it, loss=0.423]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 900): 0.3108686701953411\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 38%|███▊ | 999/2601 [24:57<39:59, 1.50s/it, loss=0.362]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1000): 0.355789770334959\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 42%|████▏ | 1099/2601 [27:26<37:32, 1.50s/it, loss=0.39] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1100): 0.3256390055269003\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 46%|████▌ | 1199/2601 [29:56<35:05, 1.50s/it, loss=0.194]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1200): 0.32589283064007757\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 50%|████▉ | 1299/2601 [32:26<32:32, 1.50s/it, loss=0.374] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1300): 0.3462872489541769\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 54%|█████▍ | 1399/2601 [34:56<30:08, 1.50s/it, loss=0.299]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1400): 0.3290054628252983\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 58%|█████▊ | 1499/2601 [37:26<27:36, 1.50s/it, loss=0.216]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1500): 0.30920374870300293\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 61%|██████▏ | 1599/2601 [39:56<25:04, 1.50s/it, loss=0.364]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1600): 0.31953610375523567\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 65%|██████▌ | 1699/2601 [42:26<22:35, 1.50s/it, loss=0.185]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1700): 0.3104771442711353\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 69%|██████▉ | 1799/2601 [44:57<20:05, 1.50s/it, loss=0.263]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1800): 0.3241106171905994\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 73%|███████▎ | 1899/2601 [47:27<17:31, 1.50s/it, loss=0.191]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 1900): 0.3105654291808605\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 77%|███████▋ | 1999/2601 [49:57<15:02, 1.50s/it, loss=0.279] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2000): 0.33756139226257803\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 81%|████████ | 2099/2601 [52:27<12:27, 1.49s/it, loss=0.388] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2100): 0.33014354825019837\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 85%|████████▍ | 2199/2601 [54:56<10:02, 1.50s/it, loss=0.302] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2200): 0.3082833808660507\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 88%|████████▊ | 2299/2601 [57:26<07:32, 1.50s/it, loss=0.19] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2300): 0.321821516752243\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 92%|█████████▏| 2399/2601 [59:56<05:02, 1.50s/it, loss=0.294] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2400): 0.32693059600889685\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 96%|█████████▌| 2499/2601 [1:02:26<02:32, 1.50s/it, loss=0.188]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2500): 0.3449856662750244\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|█████████▉| 2599/2601 [1:04:56<00:02, 1.50s/it, loss=0.201] " ] }, { "name": "stdout", "output_type": "stream", "text": [ "Avg loss for last 100 steps (step 2600): 0.3208476223796606\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Training: 100%|██████████| 2601/2601 [1:04:58<00:00, 1.50s/it, loss=0.647]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Training Loss for Epoch 5: 0.3262371334495429\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Validation: 100%|██████████| 137/137 [01:12<00:00, 1.90it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Average Validation Loss for Epoch 5: 1.1965753805898403\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "train_model(model, train_loader, val_loader)" ] }, { "cell_type": "markdown", "metadata": { "id": "pzpP5N5_CRmA" }, "source": [ "#### Saving model" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:14:12.680953Z", "iopub.status.busy": "2024-10-20T11:14:12.680601Z", "iopub.status.idle": "2024-10-20T11:14:13.827992Z", "shell.execute_reply": "2024-10-20T11:14:13.826938Z", "shell.execute_reply.started": "2024-10-20T11:14:12.680918Z" }, "id": "7MnkV0CtCRmA", "outputId": "744a3ed8-f243-4f80-bad4-71c873644901" }, "outputs": [ { "data": { "text/plain": [ "('squad-bert-trained/BERT_model/tokenizer_config.json',\n", " 'squad-bert-trained/BERT_model/special_tokens_map.json',\n", " 'squad-bert-trained/BERT_model/vocab.txt',\n", " 'squad-bert-trained/BERT_model/added_tokens.json',\n", " 'squad-bert-trained/BERT_model/tokenizer.json')" ] }, "execution_count": 45, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Define the directory where you want to save the model\n", "model_save_path = 'squad-bert-trained/BERT_model'\n", "\n", "# Create the directory if it doesn't exist\n", "if not os.path.exists(model_save_path):\n", " os.makedirs(model_save_path)\n", "\n", "\n", "# Save the trained model and tokenizer\n", "model.save_pretrained(model_save_path)\n", "tokenizer.save_pretrained(model_save_path)" ] }, { "cell_type": "markdown", "metadata": { "id": "uBRSQrYaCRmA" }, "source": [ "#### Creating a .zip folder to load in local system" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:18:15.398687Z", "iopub.status.busy": "2024-10-20T11:18:15.398050Z", "iopub.status.idle": "2024-10-20T11:18:38.861353Z", "shell.execute_reply": "2024-10-20T11:18:38.860282Z", "shell.execute_reply.started": "2024-10-20T11:18:15.398639Z" }, "id": "kyIuj_TiCRmA", "outputId": "d936a1df-ade4-4ac0-eac9-870ad0c0d672" }, "outputs": [ { "name": "stderr", "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" ] }, { "name": "stdout", "output_type": "stream", "text": [ " adding: squad-bert-trained/BERT_model/ (stored 0%)\n", " adding: squad-bert-trained/BERT_model/tokenizer_config.json (deflated 76%)\n", " adding: squad-bert-trained/BERT_model/model.safetensors (deflated 7%)\n", " adding: squad-bert-trained/BERT_model/config.json (deflated 47%)\n", " adding: squad-bert-trained/BERT_model/special_tokens_map.json (deflated 42%)\n", " adding: squad-bert-trained/BERT_model/tokenizer.json (deflated 71%)\n", " adding: squad-bert-trained/BERT_model/vocab.txt (deflated 53%)\n" ] }, { "data": { "text/html": [ "model.zip
" ], "text/plain": [ "/kaggle/working/model.zip" ] }, "execution_count": 46, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Compress the model directory into a zip file\n", "!zip -r model.zip squad-bert-trained/BERT_model\n", "\n", "# Download the zip file to your local system\n", "from IPython.display import FileLink\n", "\n", "# Display a link to download the file\n", "FileLink(r'model.zip')\n" ] }, { "cell_type": "markdown", "metadata": { "id": "Drb4Uto-CRmB" }, "source": [ "#### Loading the model and setting it to evaluation model" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:32:43.467717Z", "iopub.status.busy": "2024-10-20T11:32:43.467314Z", "iopub.status.idle": "2024-10-20T11:32:43.749348Z", "shell.execute_reply": "2024-10-20T11:32:43.748466Z", "shell.execute_reply.started": "2024-10-20T11:32:43.467677Z" }, "id": "MZbsLvxyCRmB", "outputId": "b50ba0e5-8b3d-4a11-9953-687181e892a1" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Moving the model to cuda\n", "\n", "\n" ] }, { "data": { "text/plain": [ "BertForQuestionAnswering(\n", " (bert): BertModel(\n", " (embeddings): BertEmbeddings(\n", " (word_embeddings): Embedding(30522, 768, padding_idx=0)\n", " (position_embeddings): Embedding(512, 768)\n", " (token_type_embeddings): Embedding(2, 768)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " (encoder): BertEncoder(\n", " (layer): ModuleList(\n", " (0-11): 12 x BertLayer(\n", " (attention): BertAttention(\n", " (self): BertSdpaSelfAttention(\n", " (query): Linear(in_features=768, out_features=768, bias=True)\n", " (key): Linear(in_features=768, out_features=768, bias=True)\n", " (value): Linear(in_features=768, out_features=768, bias=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " (output): BertSelfOutput(\n", " (dense): Linear(in_features=768, out_features=768, bias=True)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " )\n", " (intermediate): BertIntermediate(\n", " (dense): Linear(in_features=768, out_features=3072, bias=True)\n", " (intermediate_act_fn): GELUActivation()\n", " )\n", " (output): BertOutput(\n", " (dense): Linear(in_features=3072, out_features=768, bias=True)\n", " (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n", " (dropout): Dropout(p=0.1, inplace=False)\n", " )\n", " )\n", " )\n", " )\n", " )\n", " (qa_outputs): Linear(in_features=768, out_features=2, bias=True)\n", ")" ] }, "execution_count": 47, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Load the saved model and tokenizer\n", "model_save_path = 'squad-bert-trained/BERT_model'\n", "model = BertForQuestionAnswering.from_pretrained(model_save_path)\n", "tokenizer = BertTokenizerFast.from_pretrained(model_save_path)\n", "\n", "print(f\"Moving the model to {device}\\n\\n\")\n", "model.to(device)\n", "\n", "# Set the model to evaluation mode\n", "model.eval()" ] }, { "cell_type": "markdown", "metadata": { "id": "vPFgF0lnCRmB" }, "source": [ "**Post-Code Explanation**:\n", "The core of the training process is handled within the loop, where the model’s parameters are iteratively updated based on the training data and loss function.\n", "\n", "- **Gradient Updates**:\n", " - At the beginning of each batch iteration, `optimizer.zero_grad()` clears out the gradients from the previous iteration. This is essential because PyTorch accumulates gradients, and not clearing them would result in incorrect gradient updates.\n", " - `loss_value.backward()` performs backpropagation, calculating the gradients for the current batch of data with respect to the loss. The gradients are stored in the model's parameters.\n", " - `optimizer.step()` updates the model's weights using the gradients computed during backpropagation, adjusting the model to minimize the loss.\n", "\n", "- **Learning Rate Scheduling**:\n", " - After each batch, the learning rate scheduler (`scheduler.step()`, if included) is triggered, adjusting the learning rate dynamically based on predefined rules (such as decay). This helps in ensuring steady convergence and avoiding overshooting the optimal weights.\n", "\n", "- **Loss Tracking**:\n", " - Throughout the training process, we monitor the loss at regular intervals (e.g., every 100 batches). This involves keeping track of batch losses and printing the average, which gives us an indication of how well the model is learning over time. This step is crucial for diagnosing issues such as vanishing gradients or exploding losses early in the training process.\n", "\n", "- **Validation Phase**:\n", " - Once an epoch is completed, the model switches to evaluation mode using `model.eval()`. This phase disables dropout and batch normalization layers, which are only active during training. Additionally, `torch.no_grad()` ensures that no gradients are calculated during validation, which reduces computational overhead and memory usage.\n", " - During validation, the model's performance is assessed on unseen validation data by computing the validation loss. This serves as a proxy for how well the model generalizes to data outside the training set.\n", "\n", "- **Epoch Summary**:\n", " - At the end of each epoch, both training and validation losses are printed. These metrics are critical in diagnosing potential overfitting (training loss drops while validation loss increases) or underfitting (both losses remain high). If overfitting is detected, strategies such as early stopping or regularization might be applied in subsequent training runs.\n", "\n", "- **Fine-tuning**:\n", " - Monitoring these metrics allows us to decide if further fine-tuning is necessary. Fine-tuning might include adjustments in learning rate, model architecture, batch size, or the number of epochs. This step helps ensure the model is trained effectively while avoiding problems like overfitting or underfitting.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "tLjeYuLwCRmB" }, "source": [ "### 4. Evaluation and Inference\n", "\n", "After completing the training phase and saving the model in the previous section, we now move on to the evaluation and inference process. This section is critical in determining how well the model performs on unseen data, and it provides insights into the model's generalization capabilities.\n", "\n", "- **Inference**:\n", " - To begin, we will load the saved model checkpoint and run a few sample inputs through it to observe its predictions. This allows us to visually assess how well the model handles the task, whether it's generating coherent responses or correctly answering questions. Running individual examples also helps in detecting any obvious issues, such as grammatical errors, incompleteness, or irrelevant outputs.\n", " - We’ll also compare the model's predictions against the ground truth to evaluate its performance qualitatively.\n", "\n", "- **Evaluation Using Metrics**:\n", " - After running some basic examples, we will systematically evaluate the model using standard metrics for language models. In this project, we are using **ROUGE**, **BLEU**, and **F1** scores, which are commonly used for evaluating the quality of text generation tasks such as question answering, summarization, or machine translation.\n", " - **ROUGE (Recall-Oriented Understudy for Gisting Evaluation)**: This metric measures the overlap between the generated text and the reference text. ROUGE-1 measures the overlap of unigrams, ROUGE-2 for bigrams, and ROUGE-L for the longest common subsequence. It is especially useful for tasks like summarization, where capturing important words and phrases is critical.\n", " - **BLEU (Bilingual Evaluation Understudy)**: BLEU measures the precision of n-grams between the generated and reference texts, primarily focusing on fluency and syntactic quality. It is widely used for translation tasks.\n", " - **F1 Accuracy**: This metric evaluates the balance between precision and recall, focusing on how well the model's answers match the reference answers. It is especially important for tasks like question answering, where both exact matches and partially correct answers are considered.\n", " \n", "- **Automated Evaluation**:\n", " - We will loop through the test dataset and generate predictions for each input. For every prediction, we will calculate the ROUGE, BLEU, and F1 scores. These metrics will give us a quantitative evaluation of the model’s performance across the entire test set.\n", " - By averaging these scores, we will determine the overall performance of the model. A higher average score in all metrics indicates that the model has learned to generate meaningful, accurate, and fluent responses.\n", "\n", "- **Analysis of Results**:\n", " - Once the evaluation is complete, we will analyze the results. If the scores indicate that the model performs well (e.g., high ROUGE and BLEU scores), it means that the model has effectively learned the task and can generalize to unseen data. On the other hand, if the evaluation scores are low, it may be a sign that the model is either underfitting or overfitting.\n", " - In case of suboptimal performance, further fine-tuning of the model or hyperparameters may be required, or additional training data might be necessary to improve the results.\n", "\n", "- **Next Steps**:\n", " - Based on the evaluation results, we might consider improvements, such as adjusting the training process, refining the model’s architecture, or experimenting with data augmentation techniques. Evaluating the error cases in detail will guide future decisions to enhance the model’s overall performance.\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:33:08.831479Z", "iopub.status.busy": "2024-10-20T11:33:08.831085Z", "iopub.status.idle": "2024-10-20T11:33:08.840458Z", "shell.execute_reply": "2024-10-20T11:33:08.839365Z", "shell.execute_reply.started": "2024-10-20T11:33:08.831442Z" }, "id": "fl2R3ze6CRmB" }, "outputs": [], "source": [ "def answer_question(question, context):\n", " max_context_size = 512\n", " chunk_size = max_context_size\n", "\n", " chunks = [context[i:i+chunk_size] for i in range(0, len(context), chunk_size)]\n", "\n", " answers = []\n", " for chunk in chunks:\n", " # Tokenize the chunk\n", " inputs = tokenizer(chunk, question, return_tensors='pt', truncation=True, max_length=max_context_size).to(device)\n", "\n", " # Generate the output\n", " with torch.no_grad():\n", " outputs = model(**inputs)\n", "\n", " # Get most likely beginning and end of the answer span\n", " answer_start_scores = outputs.start_logits\n", " answer_end_scores = outputs.end_logits\n", "\n", " # Find the tokens with the highest `start` and `end` scores\n", " answer_start = torch.argmax(answer_start_scores)\n", " answer_end = torch.argmax(answer_end_scores) + 1\n", "\n", " # Convert the tokens to text\n", " answer = tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(inputs['input_ids'][0][answer_start:answer_end]))\n", "\n", " answers.append(answer)\n", "\n", " # Combine the answers from each chunk\n", " answer = ' '.join(answers)\n", " answer = answer.replace('[CLS]', '')\n", " return answer.strip()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:42:47.083484Z", "iopub.status.busy": "2024-10-20T11:42:47.083069Z", "iopub.status.idle": "2024-10-20T11:42:47.192611Z", "shell.execute_reply": "2024-10-20T11:42:47.191665Z", "shell.execute_reply.started": "2024-10-20T11:42:47.083444Z" }, "id": "whlIg4zRCRmB", "outputId": "692f1f9c-d19f-43ce-8477-291e758e2965" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Q1: When was it discovered Beyonce was a co-owner of the music service, Tidal\n", "A1: march 30, 2015 jay - z on the release of tidal.\n", "\n", "Q2: The parent company of Tidal became under the ownership of whom in 2015?\n", "A2: jay z spotify\n", "\n", "Q3: What kind of service is Tidal?\n", "A3: music streaming all artist owned\n", "\n" ] } ], "source": [ "context = \"\"\"On March 30, 2015, it was announced that Beyoncé is a co-owner, with various other music artists, in the music streaming service Tidal. The service specialises in lossless audio and high\n", "definition music videos. Beyoncé's husband Jay Z acquired the parent company of Tidal, Aspiro, in the first quarter of 2015. Including Beyoncé and Jay-Z, sixteen artist stakeholders\n", "(such as Kanye West, Rihanna, Madonna, Chris Martin, Nicki Minaj and more) co-own Tidal, with the majority owning a 3% equity stake. The idea of having an all artist owned streaming\n", "service was created by those involved to adapt to the increased demand for streaming within the current music industry, and to rival other streaming services such as Spotify, which have\n", "been criticised for their low payout of royalties. \"The challenge is to get everyone to respect music again, to recognize its value\", stated Jay-Z on the release of Tidal.\"\"\"\n", "\n", "questions = [\n", " \"When was it discovered Beyonce was a co-owner of the music service, Tidal\",\n", " \"The parent company of Tidal became under the ownership of whom in 2015?\",\n", " \"What kind of service is Tidal?\",\n", "\n", "]\n", "\n", "# True Answers for validation (if needed)\n", "True_Answers = [\"September 1876\", \"twice\", \"The Observer\", \"three\", \"1987\"]\n", "\n", "# Assuming you have a function called answer_question that returns answers based on the context and question.\n", "# Here we would loop through questions and print the results in the specified format.\n", "\n", "for index, question in enumerate(questions, start=1):\n", " answer = answer_question(question, context)\n", " print(f\"Q{index}: {question}\")\n", " print(f\"A{index}: {answer}\")\n", " print()\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:43:37.270767Z", "iopub.status.busy": "2024-10-20T11:43:37.270373Z", "iopub.status.idle": "2024-10-20T11:43:37.292985Z", "shell.execute_reply": "2024-10-20T11:43:37.292164Z", "shell.execute_reply.started": "2024-10-20T11:43:37.270730Z" }, "id": "4RNn3AFLCRmB", "outputId": "78068e75-d902-4842-ba54-89668dbcbf61" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Question: What is the difference between Music and Harmony?\n", "Answer: it is a universal language that can evoke emotions, convey messages, and bring people together. music has been an integral part of human culture for centuries, with various genres and styles emerging across the world. from classical to contemporary, music has the power to inspire, heal, and entertain\n" ] } ], "source": [ "# Test with a new question and context\n", "context = \"\"\"\n", "Music is an art form whose medium is sound and silence. It is a universal language that\n", "can evoke emotions, convey messages, and bring people together. Music has been an\n", "integral part of human culture for centuries, with various genres and styles emerging\n", "across the world. From classical to contemporary, music has the power to inspire, heal,\n", "and entertain.\n", "\"\"\"\n", "question = \"What is the difference between Music and Harmony?\"\n", "\n", "# Get the answer from the model\n", "answer = answer_question(question, context)\n", "print(f\"Question: {question}\")\n", "print(f\"Answer: {answer}\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T11:43:57.547784Z", "iopub.status.busy": "2024-10-20T11:43:57.547404Z", "iopub.status.idle": "2024-10-20T11:43:57.608202Z", "shell.execute_reply": "2024-10-20T11:43:57.607381Z", "shell.execute_reply.started": "2024-10-20T11:43:57.547745Z" }, "id": "t_XeFyKECRmC", "outputId": "d88b3671-2622-4cb7-8060-ef25b866b8e8" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Question: What are the applications of Space Robotics?\n", "Answer: space exploration, satellite maintenance, and planetary surface operations\n" ] } ], "source": [ "context = \"\"\"\n", "Space robotics is a field of robotics that focuses on the design, development, and operation of robots that can\n", "survive and function in the harsh environment of space. These robots are designed to perform a variety of tasks,\n", "such as space exploration, satellite maintenance, and planetary surface operations.\n", "Space robots can be autonomous or remotely controlled, and they often require specialized\n", "systems to withstand the extreme temperatures, radiation, and vacuum of space.\n", "\n", "Space robotics has many applications, including:\n", "\n", "Planetary exploration: Robots like NASA's Curiosity Rover and Perseverance Rover\n", "have been used to explore the surface of Mars and gather data about the planet's geology and climate.\n", "Satellite maintenance: Robots like the Canadarm2 robotic arm on the\n", "International Space Station have been used to perform maintenance tasks and repairs on satellites in orbit.\n", "Asteroid mining: Robots are being developed to explore and mine asteroids\n", "for valuable resources like water and precious metals.\n", "\"\"\"\n", "\n", "question = \"What are the applications of Space Robotics?\"\n", "\n", "# Get the answer from the model\n", "answer = answer_question(question, context)\n", "print(f\"Question: {question}\")\n", "print(f\"Answer: {answer}\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T12:06:51.353291Z", "iopub.status.busy": "2024-10-20T12:06:51.352590Z", "iopub.status.idle": "2024-10-20T12:06:51.365351Z", "shell.execute_reply": "2024-10-20T12:06:51.364392Z", "shell.execute_reply.started": "2024-10-20T12:06:51.353249Z" }, "id": "OAaexydpCRmC" }, "outputs": [], "source": [ "\n", "# Define your function for calculating metrics and generating answers\n", "def calculate_metrics(val_df):\n", " metrics = {\n", " 'rouge1': [],\n", " 'rouge2': [],\n", " 'rougel': [],\n", " 'bleu': [],\n", " 'f1': [] # Add F1 score to metrics\n", " }\n", "\n", " # Initialize the scorer\n", " scorer = rouge_scorer.RougeScorer(['rouge1', 'rouge2', 'rougeL'], use_stemmer=True)\n", "\n", " # Initialize lists to store answers and predicted answers\n", " answers = []\n", " predicted_answers = []\n", "\n", " # Iterate over the predictions\n", " for question_id in range(val_df.shape[0]):\n", " predicted_answer = answer_question(val_df.iloc[question_id]['Question'], val_df.iloc[question_id]['Context']).lower().strip()\n", " answer = val_df.iloc[question_id]['Answer'].lower().strip()\n", "\n", " # Store the answers and predicted answers\n", " answers.append(answer)\n", " predicted_answers.append(predicted_answer)\n", "\n", " # Calculate ROUGE scores\n", " rouge_scores = scorer.score(answer, predicted_answer)\n", " metrics['rouge1'].append(rouge_scores['rouge1'].fmeasure)\n", " metrics['rouge2'].append(rouge_scores['rouge2'].fmeasure)\n", " metrics['rougel'].append(rouge_scores['rougeL'].fmeasure)\n", "\n", " # Calculate BLEU score\n", " metrics['bleu'].append(sentence_bleu([answer.split()], predicted_answer.split(), weights=(1.0, 0.0, 0.0, 0.0)))\n", "\n", " # Calculate F1 score\n", " # Use a basic precision and recall approach based on the ROUGE scores\n", " precision = rouge_scores['rouge1'].precision\n", " recall = rouge_scores['rouge1'].recall\n", " if precision + recall > 0: # Avoid division by zero\n", " f1 = 2 * (precision * recall) / (precision + recall)\n", " else:\n", " f1 = 0.0\n", " metrics['f1'].append(f1)\n", "\n", " return metrics, answers, predicted_answers\n", "\n", "\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T12:06:52.671717Z", "iopub.status.busy": "2024-10-20T12:06:52.670887Z", "iopub.status.idle": "2024-10-20T12:08:23.146385Z", "shell.execute_reply": "2024-10-20T12:08:23.145359Z", "shell.execute_reply.started": "2024-10-20T12:06:52.671664Z" }, "id": "ZtT1_cIKCRmC", "outputId": "06e5029b-28be-4b41-a4db-3a4f6c16b7a7" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Average ROUGE-1: 0.5453534984095015\n", "Average ROUGE-2: 0.3224320844012193\n", "Average ROUGE-L: 0.5431188724586591\n", "Average BLEU: 0.39141109481540254\n", "Average F1 Score: 0.5453534984095015\n" ] } ], "source": [ "# Import Gaussian filter for smoothing\n", "from scipy.ndimage import gaussian_filter1d\n", "\n", "warnings.filterwarnings(\"ignore\")\n", "\n", "# Calculate metrics\n", "metrics, answers, predicted_answers = calculate_metrics(val_df)\n", "\n", "# Calculate average scores\n", "average_metrics = {key: sum(value) / len(value) for key, value in metrics.items()}\n", "\n", "# Print the average metrics\n", "print('Average ROUGE-1:', average_metrics['rouge1'])\n", "print('Average ROUGE-2:', average_metrics['rouge2'])\n", "print('Average ROUGE-L:', average_metrics['rougel'])\n", "print('Average BLEU:', average_metrics['bleu'])\n", "print('Average F1 Score:', average_metrics['f1']) # Print the average F1 score\n", "\n", "# Calculate the lengths of answers\n", "actual_lengths = [len(answer) for answer in answers]\n", "predicted_lengths = [len(predicted_answer) for predicted_answer in predicted_answers]\n", "\n", "# Create a DataFrame for easier plotting\n", "import pandas as pd\n", "\n", "lengths_df = pd.DataFrame({\n", " 'Index': range(len(actual_lengths)),\n", " 'Actual Lengths': actual_lengths,\n", " 'Predicted Lengths': predicted_lengths\n", "})\n", "\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T12:11:02.216132Z", "iopub.status.busy": "2024-10-20T12:11:02.215447Z", "iopub.status.idle": "2024-10-20T12:11:02.529747Z", "shell.execute_reply": "2024-10-20T12:11:02.528455Z", "shell.execute_reply.started": "2024-10-20T12:11:02.216090Z" }, "id": "hEjlBDrRCRmD", "outputId": "8c4bb27d-854b-4c8a-a229-2c1b22032d7b" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Scores and labels\n", "scores = [\n", " average_metrics['bleu'],\n", " average_metrics['rouge1'],\n", " average_metrics['rouge2'],\n", " average_metrics['rougel'],\n", " average_metrics['f1']\n", "]\n", "labels = ['BLEU', 'ROUGE-1', 'ROUGE-2', 'ROUGE-L', 'F1 Score']\n", "\n", "# Set the style for seaborn\n", "sns.set(style=\"darkgrid\")\n", "\n", "# Create the bar plot\n", "plt.figure(figsize=(10, 6))\n", "bar_plot = sns.barplot(x=labels, y=scores, palette= 'viridis', linewidth=1.5)\n", "\n", "# Set the width of the bars\n", "for bar in bar_plot.patches:\n", " bar.set_width(0.4) # Set the width of the bars\n", "\n", "# Adding scores on top of each bar\n", "for index, value in enumerate(scores):\n", " bar_plot.text(index, value + 0.01, f'{value:.2f}', ha='center', va='bottom')\n", "\n", "# Adding title and labels\n", "plt.title('Evaluation Metrics')\n", "plt.xlabel('Metric')\n", "plt.ylabel('Score')\n", "\n", "# Display the plot\n", "plt.show()\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "execution": { "iopub.execute_input": "2024-10-20T12:15:40.798067Z", "iopub.status.busy": "2024-10-20T12:15:40.797655Z", "iopub.status.idle": "2024-10-20T12:17:12.800263Z", "shell.execute_reply": "2024-10-20T12:17:12.799371Z", "shell.execute_reply.started": "2024-10-20T12:15:40.798030Z" }, "id": "XB-YOUqlCRmD", "outputId": "3a48cb0f-97c1-4baa-f580-e47340e287dd" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Average ROUGE-1: 0.5453534984095015\n", "Average ROUGE-2: 0.3224320844012193\n", "Average ROUGE-L: 0.5431188724586591\n", "Average BLEU: 0.39141109481540254\n", "Average F1 Score: 0.5453534984095015\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Print the average metrics\n", "print('Average ROUGE-1:', average_metrics['rouge1'])\n", "print('Average ROUGE-2:', average_metrics['rouge2'])\n", "print('Average ROUGE-L:', average_metrics['rougel'])\n", "print('Average BLEU:', average_metrics['bleu'])\n", "print('Average F1 Score:', average_metrics['f1']) # Print the average F1 score\n", "\n", "# Prepare the DataFrame for plotting\n", "metrics_df = pd.DataFrame({\n", " 'Index': range(len(metrics['rouge1'])), # Assuming all metrics have the same length\n", " 'ROUGE-1': metrics['rouge1'],\n", " 'ROUGE-2': metrics['rouge2'],\n", " 'ROUGE-L': metrics['rougel'],\n", " 'BLEU': metrics['bleu'],\n", " 'F1 Score': metrics['f1']\n", "})\n", "\n", "# Smooth the metrics using a Gaussian filter\n", "smoothed_metrics_df = pd.DataFrame({\n", " 'Index': metrics_df['Index'],\n", " 'ROUGE-1': gaussian_filter1d(metrics_df['ROUGE-1'], sigma=2),\n", " 'ROUGE-2': gaussian_filter1d(metrics_df['ROUGE-2'], sigma=2),\n", " 'ROUGE-L': gaussian_filter1d(metrics_df['ROUGE-L'], sigma=2),\n", " 'BLEU': gaussian_filter1d(metrics_df['BLEU'], sigma=2),\n", " 'F1 Score': gaussian_filter1d(metrics_df['F1 Score'], sigma=2)\n", "})\n", "\n", "# Set the style for seaborn\n", "sns.set(style=\"darkgrid\")\n", "\n", "# Create subplots for each metric\n", "fig, axs = plt.subplots(3, 2, figsize=(15, 12), constrained_layout=True)\n", "\n", "# Plot each metric\n", "axs[0, 0].plot(smoothed_metrics_df['Index'], smoothed_metrics_df['ROUGE-1'], color='blue', linewidth=2)\n", "axs[0, 0].set_title('Smoothed ROUGE-1')\n", "axs[0, 0].set_xlabel('Sample Index')\n", "axs[0, 0].set_ylabel('Score')\n", "axs[0, 0].grid()\n", "\n", "axs[0, 1].plot(smoothed_metrics_df['Index'], smoothed_metrics_df['ROUGE-2'], color='orange', linewidth=2)\n", "axs[0, 1].set_title('Smoothed ROUGE-2')\n", "axs[0, 1].set_xlabel('Sample Index')\n", "axs[0, 1].set_ylabel('Score')\n", "axs[0, 1].grid()\n", "\n", "axs[1, 0].plot(smoothed_metrics_df['Index'], smoothed_metrics_df['ROUGE-L'], color='green', linewidth=2)\n", "axs[1, 0].set_title('Smoothed ROUGE-L')\n", "axs[1, 0].set_xlabel('Sample Index')\n", "axs[1, 0].set_ylabel('Score')\n", "axs[1, 0].grid()\n", "\n", "axs[1, 1].plot(smoothed_metrics_df['Index'], smoothed_metrics_df['BLEU'], color='red', linewidth=2)\n", "axs[1, 1].set_title('Smoothed BLEU Score')\n", "axs[1, 1].set_xlabel('Sample Index')\n", "axs[1, 1].set_ylabel('Score')\n", "axs[1, 1].grid()\n", "\n", "axs[2, 0].plot(smoothed_metrics_df['Index'], smoothed_metrics_df['F1 Score'], color='purple', linewidth=2)\n", "axs[2, 0].set_title('Smoothed F1 Score')\n", "axs[2, 0].set_xlabel('Sample Index')\n", "axs[2, 0].set_ylabel('Score')\n", "axs[2, 0].grid()\n", "\n", "# Remove the empty subplot\n", "fig.delaxes(axs[2, 1])\n", "\n", "# Show the plots\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": { "id": "s2uxqxCHCRmD" }, "source": [ "This code defines a system for evaluating question-answering models based on various metrics, such as ROUGE, BLEU, and F1 score. The `calculate_metrics` function processes a dataset of questions, contexts, and answers to generate predicted answers using a pre-trained model. It computes ROUGE-1, ROUGE-2, ROUGE-L, and BLEU scores for each prediction, as well as an F1 score using precision and recall from the ROUGE-1 score.\n", "\n", "The code also smooths the metric values using a Gaussian filter, which helps to visualize trends more clearly. It plots the smoothed ROUGE-1, ROUGE-2, ROUGE-L, BLEU, and F1 scores in subplots to track the model's performance across different samples. The final metrics and smoothed visualizations provide insights into how well the model answers questions relative to the ground truth, helping to assess its accuracy and effectiveness in the task.\n", "\n", "This code evaluates the performance of a question-answering model using various metrics to assess its accuracy and effectiveness. It includes two primary functions: `calculate_metrics` and a visualization section.\n", "\n", "1. **`calculate_metrics(val_df)`**:\n", " - This function takes a validation DataFrame (`val_df`) containing questions, contexts, and true answers.\n", " - It initializes a dictionary to store ROUGE-1, ROUGE-2, ROUGE-L, BLEU, and F1 scores.\n", " - Using the `RougeScorer`, it computes the ROUGE scores for each predicted answer against the true answer. For each entry, it:\n", " - Calls the `answer_question` function to generate a predicted answer.\n", " - Appends the computed ROUGE scores, BLEU score (based on the predicted and true answers), and F1 score (calculated using precision and recall from ROUGE-1) to the metrics dictionary.\n", " - The function returns the complete metrics, along with lists of true answers and predicted answers.\n", "\n", "2. **Metric Calculation and Averaging**:\n", " - After calculating metrics for all questions, the code computes the average values for ROUGE-1, ROUGE-2, ROUGE-L, BLEU, and F1 score to provide an overall assessment of the model's performance.\n", "\n", "3. **Length Calculation**:\n", " - The code calculates the lengths of the actual and predicted answers, storing these lengths in lists for further analysis.\n", "\n", "4. **Visualization**:\n", " - It prepares a DataFrame (`lengths_df`) to facilitate plotting the lengths of actual and predicted answers.\n", " - Finally, it creates a separate DataFrame (`metrics_df`) for visualizing the individual metric scores over different samples, applying a Gaussian filter to smooth the results.\n", " - The code generates subplots for each metric (ROUGE-1, ROUGE-2, ROUGE-L, BLEU, and F1 score), allowing for a visual assessment of the model's performance trends.\n", "\n", "Overall, this evaluation framework enables an in-depth analysis of how well the question-answering model performs across various metrics, providing insights into its strengths and weaknesses.\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "gDIL9xHlCRmD" }, "source": [ "### Inference on Model Evaluation Results\n", "\n", "The evaluation metrics we’ve gathered shed light on how well our question-answering model is performing, especially when it comes to understanding context and generating responses. Here’s a closer look at what these numbers mean and how they might impact real-world business applications:\n", "\n", "1. **Average ROUGE-1: 0.545**:\n", " - This metric measures how many of the individual words in the model's answers overlap with those in the reference answers. A score of 0.545 indicates a decent level of agreement.\n", " - **Implication**: For businesses, this suggests that our model can pick up key information quite well. Imagine a customer support chatbot that can understand and respond to common inquiries accurately—that’s what this score implies. Customers would likely receive relevant answers to their questions, improving their experience.\n", "\n", "2. **Average ROUGE-2: 0.322**:\n", " - ROUGE-2 looks at pairs of consecutive words. A score of 0.322 indicates that the model struggles a bit more here.\n", " - **Implication**: This tells us that while our model is good at picking out individual words, it sometimes misses the mark when it comes to understanding phrases or context. In a business setting, this could impact things like product descriptions or service explanations, where precise language is essential for clarity and customer understanding.\n", "\n", "3. **Average ROUGE-L: 0.543**:\n", " - This metric considers the longest common subsequence of words in the answers. A score of 0.543 suggests the model maintains a fair amount of coherence in its responses.\n", " - **Implication**: This is important for user engagement. Think about virtual assistants that need to provide detailed answers—this score indicates that the model can create responses that flow reasonably well, making it easier for users to follow along.\n", "\n", "4. **Average BLEU: 0.391**:\n", " - BLEU measures how well the generated answers match the expected responses, focusing on precision. An average score of 0.391 indicates that while the answers are relevant, they may not always capture the full context perfectly.\n", " - **Implication**: In practical terms, this could affect applications like content generation or translation. Businesses might find that while the model can generate helpful content, there’s still room for improvement. This could mean more editing or refining to ensure the final product meets quality standards.\n", "\n", "5. **Average F1 Score: 0.545**:\n", " - The F1 score balances precision (how many of the returned answers were correct) and recall (how many correct answers were returned). A score of 0.545 suggests moderate performance.\n", " - **Implication**: For businesses, this score indicates that the model is fairly good at providing accurate responses without overwhelming users with incorrect information. This balance is particularly important in sensitive fields like healthcare or finance, where getting things right can make a big difference.\n", "\n", "### Real-World Business Applications\n", "- **Context-Based Customer Support**: These metrics suggest that the model could be effectively integrated into customer service platforms, automating responses to frequently asked questions based on context. This could enhance user satisfaction and reduce response times.\n", "\n", "- **Contextual Content Generation**: The model shows promise for generating contextually relevant content for blogs, reports, and marketing materials. While it performs decently, human review will be essential to align the generated content with the brand's voice and ensure it resonates with the target audience.\n", "\n", "- **Contextual Data Analysis**: Teams could leverage the model to summarize findings or generate insights from reports, enhancing productivity and allowing decision-makers to focus on crucial information without getting bogged down in details.\n", "\n", "- **Advanced Chatbots and Virtual Assistants**: The results suggest that while the model can provide useful context-based answers, further fine-tuning is needed to fully grasp user intent and context. Improving this aspect could significantly enhance user interactions, making them feel more natural and intuitive.\n", "\n", "In summary, these evaluation metrics indicate that our model has a solid foundation for context-based question answering but also highlight opportunities for enhancement. As we continue to refine its capabilities, we’re paving the way for more effective tools that can genuinely benefit businesses and their customers.\n" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "id": "6kHhDIJ0CRmD" }, "outputs": [ { "data": { "text/plain": [ "'C:\\\\Users\\\\mohds\\\\OneDrive - University of San Diego\\\\Desktop\\\\MS_AAI\\\\Natural Language Processing and GenAI (AAI-520-A1)\\\\EDITH_ChatBOT'" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pwd" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "colab": { "provenance": [] }, "kaggle": { "accelerator": "nvidiaTeslaT4", "dataSources": [], "dockerImageVersionId": 30787, "isGpuEnabled": true, "isInternetEnabled": true, "language": "python", "sourceType": "notebook" }, "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": "3.11.4" } }, "nbformat": 4, "nbformat_minor": 1 }