THaMES: An End-to-End Tool for Hallucination Mitigation and Evaluation in Large Language Models