Benchmark Evaluations, Applications, and Challenges of Large Vision Language Models: A Survey