Added multi specialization for Qwen2.5-VL, Qwen3-VL and Qwen3_VL_MOE models as per reference from #909. #1021
+887
−223
Loading