Toxic Comment Classifier

In this project, I developed a TensorFlow-based model leveraging bidirectional LSTM layers to classify internet comments across six categories: toxic, severely toxic, obscene, threat, insult, and identity hate. Using a dataset of 16,000 comments, I trained the model to accurately identify these attributes, demonstrating my proficiency in natural language processing and deep learning techniques. To make this solution accessible, I deployed it as an interactive Gradio app, allowing users to input sentences and receive real-time predictions about the comment's toxicity. This project highlights my ability to build and deploy NLP models to address real-world challenges in content moderation and online safety.