Knowing Where to Focus: Event-aware Transformer for Video Grounding