From 00c361fe53e5fc105a077f90a0a22d4c60936ffe Mon Sep 17 00:00:00 2001
From: "Gilad S." <7817232+giladgd@users.noreply.github.com>
Date: Mon, 1 Dec 2025 22:21:13 +0200
Subject: [PATCH] fix: llama arch implementation (#17665)

---
 src/llama-model.cpp | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/llama-model.cpp b/src/llama-model.cpp
index 584efbf3c8..c46ee3707f 100644
--- a/src/llama-model.cpp
+++ b/src/llama-model.cpp
@@ -626,6 +626,8 @@ void llama_model::load_hparams(llama_model_loader & ml) {
     switch (arch) {
         case LLM_ARCH_LLAMA:
             {
+                ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
+
                 if (hparams.n_expert == 8) {
                     switch (hparams.n_layer) {
                         case 32: type = LLM_TYPE_8x7B; break;