The official code implementation of paper "Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models" - View it on GitHub