models : Attach Mistral3 NVFP4 weight scales (#23629)

diff --git a/src/models/mistral3.cpp b/src/models/mistral3.cpp
index 4e6ebef82cbbd67dd3afda08c5f9cc274f34cb72..1ac5a95ccdc5a4a773cfd3adbd64745d45c436b3 100644 (file)
--- a/src/models/mistral3.cpp
+++ b/src/models/mistral3.cpp
@@ -177,9 +177,9 @@ llama_model_mistral3::graph::graph(const llama_model & model, const llm_graph_pa
             cb(cur, "ffn_norm", il);
 
             cur = build_ffn(cur,
-                    model.layers[il].ffn_up,   model.layers[il].ffn_up_b,   NULL,
-                    model.layers[il].ffn_gate, model.layers[il].ffn_gate_b, NULL,
-                    model.layers[il].ffn_down, model.layers[il].ffn_down_b, NULL,
+                    model.layers[il].ffn_up,   model.layers[il].ffn_up_b,   model.layers[il].ffn_up_s,
+                    model.layers[il].ffn_gate, model.layers[il].ffn_gate_b, model.layers[il].ffn_gate_s,
+                    model.layers[il].ffn_down, model.layers[il].ffn_down_b, model.layers[il].ffn_down_s,
                     NULL,
                     LLM_FFN_SILU, LLM_FFN_PAR, il);
             cb(cur, "ffn_out", il);
@@ -200,7 +200,11 @@ llama_model_mistral3::graph::graph(const llama_model & model, const llm_graph_pa
                     LLM_FFN_SILU, true,
                     hparams.expert_weights_scale,
                     LLAMA_EXPERT_GATING_FUNC_TYPE_SOFTMAX,
-                    il);
+                    il,
+                    nullptr, nullptr,
+                    model.layers[il].ffn_up_exps_s,
+                    model.layers[il].ffn_gate_exps_s,
+                    model.layers[il].ffn_down_exps_s);
             cb(cur, "ffn_moe_out", il);
         }
         cur = ggml_add(ctx0, cur, ffn_inp);