File size: 4,580 Bytes
2661812 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 |
# -*- coding: utf-8 -*-
"""Fake_News Generator_Detector.ipynb
Automatically generated by Colab.
Original file is located at
https://colab.research.google.com/drive/1C8ZWvJVZzTCngwpz788EpcT2lrbBWhYa
### 📰 **Project Description: Fake News Generator & Detector using Generative AI and NLP**
This project is an interactive AI-powered tool that enables users to both generate realistic-looking fake news articles and detect whether a given news text is fake or real using state-of-the-art natural language processing models. It highlights how language models can both create and combat misinformation.
🧰 **The application is built using:**
1. **Transformers Library (by Hugging Face)**:
For loading and using pretrained models — GPT-2 for fake news generation and fine-tuned BERT for fake news detection.
2. **Gradio:** To create a simple, interactive web-based user interface with tabbed sections for generation and detection.
3. **Google Colab / Python:** For backend development, prototyping, and running the application in a cloud-based notebook environment.
🎯 **Project Objectives**
**1. Generate Fake News Text**
Use GPT-2 to simulate fake news articles from user-provided prompts for for awareness and experimentation.
**2. Detect Fake or Real News**
Utilize a fine-tuned BERT model (Pulk17/Fake-News-Detection) to accurately classify news content as fake or real.
**3. Provide an Interactive Interface**
Use Gradio to build a user-friendly web interface.
**4. Demonstrate Dual Use of AI in Misinformation**
Showcase how AI can both create and detect fake news, promoting awareness and responsible AI usage.
**Step 2: Import Required Modules**
Import libraries for model handling, tokenization, and interface building.
"""
# app.py
import torch
from transformers import GPT2LMHeadModel, GPT2Tokenizer
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import gradio as gr
# Device setup
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# Load GPT-2 for fake news generation
gpt2_tokenizer = GPT2Tokenizer.from_pretrained("gpt2")
gpt2_model = GPT2LMHeadModel.from_pretrained("gpt2").to(device)
# Load fine-tuned BERT model for fake news detection
bert_tokenizer = AutoTokenizer.from_pretrained("Pulk17/Fake-News-Detection")
bert_model = AutoModelForSequenceClassification.from_pretrained(
"Pulk17/Fake-News-Detection"
).to(device)
# Function to generate fake news from a prompt
def generate_fake_news(prompt):
inputs = gpt2_tokenizer.encode(prompt, return_tensors="pt").to(device)
outputs = gpt2_model.generate(
inputs,
max_length=200,
num_return_sequences=1,
no_repeat_ngram_size=2,
do_sample=True,
temperature=0.7,
top_k=50,
top_p=0.95,
early_stopping=True
)
generated_text = gpt2_tokenizer.decode(outputs[0], skip_special_tokens=True)
return generated_text
# Function to detect if news is fake or real
def detect_news(text):
inputs = bert_tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
with torch.no_grad():
outputs = bert_model(**inputs)
logits = outputs.logits
predicted_class = torch.argmax(logits, dim=1).item()
confidence = torch.softmax(logits, dim=1)[0][predicted_class].item()
label = "🟥 Fake News" if predicted_class == 0 else "🟩 Real News"
return f"{label} (Confidence: {confidence:.2f})"
# Gradio Interface
with gr.Blocks() as demo:
gr.Markdown("## 📰 Fake News Generator & Detector (GPT-2 + BERT)")
with gr.Tab("🛠️ Generate Fake News"):
with gr.Row():
input_text = gr.Textbox(
label="Enter a News Headline or Prompt",
placeholder="e.g. Scientists discover a talking dolphin species near Japan...",
lines=2
)
generate_btn = gr.Button("Generate")
output_text = gr.Textbox(label="Generated News Article")
generate_btn.click(generate_fake_news, inputs=input_text, outputs=output_text)
with gr.Tab("🔍 Detect Fake or Real"):
with gr.Row():
detect_input = gr.Textbox(
label="Enter a News Article or Statement",
placeholder="Paste a paragraph to detect if it's fake or real...",
lines=5
)
detect_btn = gr.Button("Detect")
detect_output = gr.Textbox(label="Detection Result")
detect_btn.click(detect_news, inputs=detect_input, outputs=detect_output)
# Launch the app
demo.launch()
|