Commit 961c376a2 for llama.cpp

commit 961c376a2daa42220ec09663b9bde32c5fba1f70
Author: begcdn <150344566+begcdn@users.noreply.github.com>
Date:   Tue Oct 27 06:04:19 2026 +0800

    llama : enable moe amax rescaling for laguna (#31147)

    * llama : enable moe amax rescaling for laguna

    Assisted-by: Devin

    * llama : declare laguna sm tensor support

diff --git a/src/llama-arch.cpp b/src/llama-arch.cpp
index 06fc56bb2..fe3ad3eb7 100644
--- a/src/llama-arch.cpp
+++ b/src/llama-arch.cpp
@@ -1347,6 +1347,7 @@ bool llm_arch_supports_sm_tensor(const llm_arch & arch) {
         case LLM_ARCH_MISTRAL4:
         case LLM_ARCH_MELLUM: // [TAG_MOE_AMAX_SCALE]
         case LLM_ARCH_FLINT:  // [TAG_MOE_AMAX_SCALE]
+        case LLM_ARCH_LAGUNA: // [TAG_MOE_AMAX_SCALE]
         case LLM_ARCH_KIMI_LINEAR:
         case LLM_ARCH_BAILINGMOE3:
         case LLM_ARCH_KIMI_K3:
diff --git a/src/llama-graph.cpp b/src/llama-graph.cpp
index a22dfba96..7309705dc 100644
--- a/src/llama-graph.cpp
+++ b/src/llama-graph.cpp
@@ -2519,7 +2519,7 @@ ggml_tensor * llm_graph_context::build_moe_ffn(
         //       ref: https://github.com/ggml-org/llama.cpp/pull/30656#issuecomment-6269597922
         llm_graph_amax_src1 amax_src1 = { nullptr, nullptr };

-        if (arch == LLM_ARCH_MISTRAL4 || arch == LLM_ARCH_MELLUM || arch == LLM_ARCH_FLINT) {
+        if (arch == LLM_ARCH_MISTRAL4 || arch == LLM_ARCH_MELLUM || arch == LLM_ARCH_FLINT || arch == LLM_ARCH_LAGUNA) {
             amax_src1 = build_amax_src1_f16(cur);
             cur = amax_src1.res;
         }