nlp_trainer.py

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from transformers import AutoTokenizer, AutoModelForSequenceClassification
from transformers import Trainer, TrainingArguments
from torch.utils.data import Dataset
import torch
import re
import string
import logging
logging.basicConfig(level=logging.INFO)

def load_dataset(path="./combined.csv"):
    df = pd.read_csv(path, dtype={'text': str, 'label': str})  # Explicitly set dtypes
    df = df.dropna()  # Remove any null values
    
    # Ensure consistent column names
    if 'news' in df.columns:
        df = df.rename(columns={"news": "text"})
    if 'target' in df.columns:
        df = df.rename(columns={"target": "label"})
    
    # Convert labels to integers safely
    label_map = {"real": 0, "fake": 1}
    df['label'] = df['label'].str.lower().map(label_map)
    
    # Drop any rows where label mapping failed
    df = df.dropna(subset=['label'])
    df['label'] = df['label'].astype(int)
    
    X = df['text'].apply(str).tolist()  # Ensure text is string
    y = df['label'].tolist()
    
    return train_test_split(X, y, test_size=0.2, random_state=42)

class NewsDataset(Dataset):
    def __init__(self, texts, labels, tokenizer, max_len):
        self.texts = texts
        self.labels = labels
        self.tokenizer = tokenizer
        self.max_len = max_len

    def __len__(self):
        return len(self.texts)

    def __getitem__(self, idx):
        text = str(self.texts[idx])
        encoding = self.tokenizer(
            text,
            max_length=self.max_len,
            padding='max_length',
            truncation=True,
            return_tensors="pt"
        )
        return {
            'input_ids': encoding['input_ids'].squeeze(0),
            'attention_mask': encoding['attention_mask'].squeeze(0),
            'labels': torch.tensor(int(self.labels[idx]), dtype=torch.long)
        }

def train_model(train_texts, train_labels, val_texts, val_labels):
    tokenizer = AutoTokenizer.from_pretrained('microsoft/deberta-v3-small')
    model = AutoModelForSequenceClassification.from_pretrained('microsoft/deberta-v3-small', num_labels=2)

    train_dataset = NewsDataset(train_texts, train_labels, tokenizer, max_len=128)
    val_dataset = NewsDataset(val_texts, val_labels, tokenizer, max_len=128)

    training_args = TrainingArguments(
        output_dir='./results',
        num_train_epochs=5,
        per_device_train_batch_size=8,
        per_device_eval_batch_size=8,
        warmup_steps=500,
        weight_decay=0.01,
        logging_dir='./logs',
        evaluation_strategy="epoch",
        save_strategy="epoch"
    )

    trainer = Trainer(
        model=model,
        args=training_args,
        train_dataset=train_dataset,
        eval_dataset=val_dataset
    )

    trainer.train()
    return tokenizer, model

def predict_news(tokenizer, model, news_text):
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    model.to(device)
    print(device)
    model.eval()

    encoding = tokenizer(
        str(news_text),
        max_length=128,
        padding='max_length',
        truncation=True,
        return_tensors="pt"
    )
    
    input_ids = encoding['input_ids'].to(device)
    attention_mask = encoding['attention_mask'].to(device)

    with torch.no_grad():
        outputs = model(input_ids=input_ids, attention_mask=attention_mask)
        prediction = torch.argmax(outputs.logits, dim=1).item()

    return "Fake" if prediction == 1 else "Real"

def main():
    try:
        X_train, X_test, y_train, y_test = load_dataset()
        tokenizer, model = train_model(X_train, y_train, X_test, y_test)

        while True:
            user_input = input("\nEnter news text (or 'exit' to quit): ")
            if user_input.lower() == 'exit':
                break
            result = predict_news(tokenizer, model, user_input)
            print(f"The news is: {result}")

    except Exception as e:
        logging.error(f"An error occurred: {str(e)}")
        raise

if __name__ == "__main__":
    main()