llama: initialize pre-norm embedding mask flag (#23256)

diff --git a/src/llama-context.cpp b/src/llama-context.cpp
index b1b12d017c0a9d0a2599c871db979db2bc5e3715..3cc8ffa6668d0d09ced09c11452b1ad8a5f743b3 100644 (file)
--- a/src/llama-context.cpp
+++ b/src/llama-context.cpp
@@ -64,8 +64,9 @@ llama_context::llama_context(
     cparams.yarn_attn_factor = params.yarn_attn_factor >= 0.0f ? params.yarn_attn_factor : hparams.yarn_attn_factor;
     cparams.yarn_beta_fast   = params.yarn_beta_fast   >= 0.0f ? params.yarn_beta_fast   : hparams.yarn_beta_fast;
     cparams.yarn_beta_slow   = params.yarn_beta_slow   >= 0.0f ? params.yarn_beta_slow   : hparams.yarn_beta_slow;
-    cparams.embeddings       = params.embeddings;
-    cparams.embeddings_pre_norm = false;
+    cparams.embeddings                  = params.embeddings;
+    cparams.embeddings_pre_norm         = false;
+    cparams.embeddings_pre_norm_masked  = false;
     cparams.offload_kqv      = params.offload_kqv;
     cparams.no_perf          = params.no_perf;
     cparams.pooling_type     = params.pooling_type;