Can Large Language Models be a Cardinality Estimator? An Empirical study | ResearchPod