DistilBERT-AI-Text-Detector

About Model

The DistilBERT AI Text Detector is an experimental binary text classification model built by fine-tuning distilbert-base-uncased to distinguish between AI-generated and human-written text. Designed for sequence classification, it maps the output label 0 to human-written text and label 1 to AI-generated text. The model was trained using PyTorch and Hugging Face Transformers over 10 epochs with a batch size of 16 and a learning rate of 5e-6 on a small custom dataset containing approximately 1.4k samples. During validation, the model reached an accuracy of 0.5730 (57.3%), a precision of 0.6162, a recall of 0.9858, and an F1-score of 0.6814. To use this model, load both the model and tokenizer directly via the Hugging Face Transformers library, pass your text input, and extract the resulting classification label. Distributed under the cc-by-nc-sa-4.0 license, this model remains strictly experimental and is not intended for production environments.