Hugging Face · 每日论文·· 2 天前AI 评分32
Skill-Decoupled On-Policy Expert Distillation for Unified Vision-Language Models
Skill-Decoupled On-Policy Expert Distillation for Unified Vision-Language Models
AI 导读
Skill-Decoupled On-Policy Expert Distillation (SCOPED) 提出了一种训练统一视觉-语言模型的新方法,通过分离技能并利用奖励引导的反思机制提升模型性能。
来源:Hugging Face · 每日论文 · huggingface.co