# Open-source SOTA

LLaVA-NeXT-Video is an open-source multimodal chatbot that achieves excellent video understanding capabilities through mixed training on video and image data, reaching SOTA level among open-source models on the VideoMME benchmark.

A reward model trained on Meta-Llama-3-8B-Instruct for reward modeling in RLHF processes, supporting PPO, iterative SFT, and iterative DPO methods.