Better optimization log for gpt-oss model (#8953)

python/sglang/srt/server_args.py

@@ -488,8 +488,14 @@ class ServerArgs:
             if is_sm100_supported() and is_mxfp4_quant_format:
                 self.enable_flashinfer_mxfp4_moe = True
                 self.enable_triton_kernel_moe = False
+                logger.info(
+                    "Detected SM100 and MXFP4 quantization format for GPT-OSS model, enabling FlashInfer MXFP4 MOE kernel."
+                )
             else:
                 self.enable_triton_kernel_moe = True
+                logger.info(
+                    "Detected GPT-OSS model, enabling triton_kernels MOE kernel."
+                )
 
             self.disable_hybrid_swa_memory = True