# Multimodal Video Analysis

LLaVA-Video-7B-Qwen2-TPO is a video understanding model based on LLaVA-Video-7B-Qwen2 with temporal preference optimization, demonstrating excellent performance across multiple benchmarks.