DPC: Dual-Prompt Collaboration for Tuning Vision-Language Models | Read Paper on Bytez