Fetching the paper…

Ziya-Visual: Bilingual Large Vision-Language Model via Multi-Task Instruction Tuning · Around