models : move build_inp_out_ids outside loop (#17151)

* move build_inp_out_ids outside loop

* realign

src/models/ernie4-5.cpp

@@ -1,7 +1,5 @@
 #include "models.h"
 
-
-
 llm_build_ernie4_5::llm_build_ernie4_5(const llama_model & model, const llm_graph_params & params) :
     llm_graph_context(params) {
     const int64_t n_embd_head = hparams.n_embd_head_v;
@@ -19,6 +17,8 @@ llm_build_ernie4_5::llm_build_ernie4_5(const llama_model & model, const llm_grap
 
     auto * inp_attn = build_attn_inp_kv();
 
+    ggml_tensor * inp_out_ids = build_inp_out_ids();
+
     for (int il = 0; il < n_layer; ++il) {
         ggml_tensor * inpSA = inpL;
 
@@ -67,9 +67,8 @@ llm_build_ernie4_5::llm_build_ernie4_5(const llama_model & model, const llm_grap
         }
         if (il == n_layer - 1) {
             // skip computing output for unused tokens
-            ggml_tensor * inp_out_ids = build_inp_out_ids();
-            cur                       = ggml_get_rows(ctx0, cur, inp_out_ids);
-            inpSA                     = ggml_get_rows(ctx0, inpSA, inp_out_ids);
+            cur   = ggml_get_rows(ctx0, cur, inp_out_ids);
+            inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
         }
         ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
         cb(ffn_inp, "ffn_inp", il);

src/models/openai-moe-iswa.cpp

@@ -11,6 +11,8 @@ llm_build_openai_moe_iswa::llm_build_openai_moe_iswa(const llama_model & model,
 
     auto * inp_attn = build_attn_inp_kv_iswa();
 
+    ggml_tensor * inp_out_ids = build_inp_out_ids();
+
     for (int il = 0; il < n_layer; ++il) {
         ggml_tensor * inpSA = inpL;
 
@@ -69,7 +71,6 @@ llm_build_openai_moe_iswa::llm_build_openai_moe_iswa(const llama_model & model,
         }
         if (il == n_layer - 1) {
             // skip computing output for unused tokens
-            ggml_tensor * inp_out_ids = build_inp_out_ids();
             cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);
             inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
         }