Unlocking the power of multimodal general AI. See more recent works of our group at https://github.com/QwenLM