Bytedance_vidi2_5

Model and technical report released. Multimodal LLMs for video understanding: ViDi2.5🔍
video editing/creation: ViDi-Edit 🎬. [demo] [paper]