This repository contains the implementation of a chatbot using a Sequence-to-Sequence (Seq2Seq) model with attention mechanism. The heart of the model is the GRU (Gated Recurrent Unit) unit, which can be tailored to be either Bidirectional or Unidirectional, adapting to specific needs. The chatbot is trained to generate responses in a conversation-like manner. The code includes data preprocessing, model architecture, training, and evaluation components. It supports both Greedy Search and Beam Search for response generation, providing users with options for different response styles. The model is built in PyTorch and offers easy customization for different datasets and hyperparameters. This repository aims to provide a solid foundation for building conversational AI applications.