train.py

from model.model import initialize_model, set_seed, train, \
    evaluate, bert_predict, scoring_metrics
from model.prepare_data import load_data, score_to_sentiment,\
    preprocessing_for_bert, create_data_loaders, evaluate_roc
import torch
import numpy as np
import pandas as pd
import seaborn as sns
from pylab import rcParams
from sklearn.model_selection import train_test_split
import nltk
words = set(nltk.corpus.words.words())
sns.set(style='whitegrid', palette='muted', font_scale=1.2)
HAPPY_COLORS_PALETTE = [
    "#01BEFE",
    "#FFDD00",
    "#FF7D00",
    "#FF006D",
    "#ADFF02",
    "#8F00FF"]
sns.set_palette(sns.color_palette(HAPPY_COLORS_PALETTE))
rcParams['figure.figsize'] = 12, 8
RANDOM_SEED = 0
np.random.seed(RANDOM_SEED)
torch.manual_seed(RANDOM_SEED)
pd.set_option('display.max_columns', None)
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

if __name__ == '__main__':
    df = load_data('data/app_reviews.csv')
    df['sentiment_score'] = df['score'].apply(score_to_sentiment)
    df['sentiment'] = df['sentiment_score'].replace(
        {1: 'negative', 0: 'neutral', 2: 'positive'})
    df = df[df['sentiment'] != 'neutral']
    df['sentiment_score'] = df['sentiment_score'].replace({1: 0, 2: 1})
    df = df.dropna()
    train_content, val_content, train_sentiments, \
        val_sentiments = train_test_split(
            df['content'], df['sentiment_score'], test_size=0.2,
            random_state=RANDOM_SEED, stratify=df['sentiment'])
    train_inputs, train_masks = preprocessing_for_bert(train_content)
    val_inputs, val_masks = preprocessing_for_bert(val_content)
    train_dataloader, val_dataloader = create_data_loaders(
        train_sentiments, val_sentiments, train_masks,
        val_masks, train_inputs, val_inputs, 32)
    set_seed(42)

    bert_classifier, optimizer, scheduler = initialize_model(
        train_dataloader, epochs=8)
    train(
        bert_classifier,
        optimizer,
        scheduler,
        train_dataloader,
        val_dataloader,
        epochs=8,
        evaluation=True)
    torch.save(bert_classifier.state_dict(), 'model_weights.pt')
    path = 'model_weights.pt'
    bert_classifier.load_state_dict(torch.load(path))
    evaluate(bert_classifier, val_dataloader)
    preds = bert_predict(bert_classifier, val_dataloader)
    evaluate_roc(preds, val_sentiments)
    threshold = 0.4
    preds = np.where(preds[:, 1] > threshold, 1, 0)
    scoring_metrics(preds, val_sentiments)

    input_batch = next(iter(train_dataloader))
    input_sample = {
        "input_ids": input_batch[0][0].unsqueeze(0),
        "attention_mask": input_batch[1][0].unsqueeze(0),
    }