Transformer models have revolutionized natural language processing, achieving state-of-the-art results on a variety of tasks. However, these powerful models are susceptible to inheriting and amplifying biases present in the training data. addressing bias and achieving fairness in transformer models