AI-Generated Text Detector

This repository contains code for fine-tuning a BERT model to detect AI-generated text. The model can classify text as either 'student' generated or 'AI' generated.

Setup Instructions

Prerequisites

• Python 3.12
• Jupyter Notebook
• CUDA-enabled GPU (optional but recommended for faster training)
• Required Python packages (listed in requirements.txt)

Installation

1. Clone the repository:

   git clone https://github.com/HARSHDIPSAHA/AI-generated-text-detector.git
   cd AI-generated-text-detector

2. Create a virtual environment:

   python -m venv myenv
   source myenv/bin/activate   # On Windows, use `myenv\Scripts\activate`

3. Install the required packages:

   pip install -r requirements.txt

Running the Notebook

1. Launch Jupyter Notebook:

   jupyter notebook

2. Open berttttt.ipynb in Jupyter Notebook.

3. Run all cells to train the model and make predictions.

Using Git LFS

If you don't want to run Jupyter Notebooks, you can use Git LFS to download the pre-trained model and tokenizer:

1. Install Git LFS:

   git lfs install

2. Clone the repository with Git LFS:

   git lfs clone https://github.com/HARSHDIPSAHA/AI-generated-text-detector.git
   cd AI-generated-text-detector

3. Copy the bert_finetuned_model and bert_tokenizer directories to your local machine.

4. Create a new Jupyter Notebook and load the model:

   import torch
   from transformers import BertTokenizer, BertForSequenceClassification
   
   def load_model_and_tokenizer(model_path, tokenizer_path):
       model = BertForSequenceClassification.from_pretrained(model_path)
       tokenizer = BertTokenizer.from_pretrained(tokenizer_path)
       return model, tokenizer
   
   model_path = './bert_finetuned_model'
   tokenizer_path = './bert_tokenizer'
   model, tokenizer = load_model_and_tokenizer(model_path, tokenizer_path)
   
   labels = {0: "student", 1: "ai"}
   
   def predict_text_category(dialogue, model, tokenizer):
       inputs = tokenizer(dialogue, return_tensors='pt', truncation=True, padding=True, max_length=512)
       with torch.no_grad():
           outputs = model(**inputs)
       logits = outputs.logits
       predicted_class = torch.argmax(logits, dim=1).item()
       return labels[predicted_class]
   
   text = "Blockchain is revolutionizing education, providing personalized learning experiences to students all over the world."
   predicted_label = predict_text_category(text, model, tokenizer)
   print(f"The predicted label for the given text is: {predicted_label}")
   (1)

Example Predictions

text = "Let us eat together any burger."
predicted_label = predict_text_category(text, model, tokenizer)
print(f"The predicted label for the given text is: {predicted_label}")
(0)