diff --git a/common/common.cpp b/common/common.cpp
index 43c8840ad..e7b8ab1d1 100644
--- a/common/common.cpp
+++ b/common/common.cpp
@@ -1635,7 +1635,7 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
     options.push_back({ "server",      "       --host HOST",            "ip address to listen (default: %s)", params.hostname.c_str() });
     options.push_back({ "server",      "       --port PORT",            "port to listen (default: %d)", params.port });
     options.push_back({ "server",      "       --path PATH",            "path to serve static files from (default: %s)", params.public_path.c_str() });
-    options.push_back({ "server",      "       --embedding(s)",         "enable embedding endpoint (default: %s)", params.embedding ? "enabled" : "disabled" });
+    options.push_back({ "server",      "       --embedding(s)",         "restrict to only support embedding use case; use only with dedicated embedding models (default: %s)", params.embedding ? "enabled" : "disabled" });
     options.push_back({ "server",      "       --api-key KEY",          "API key to use for authentication (default: none)" });
     options.push_back({ "server",      "       --api-key-file FNAME",   "path to file containing API keys (default: none)" });
     options.push_back({ "server",      "       --ssl-key-file FNAME",   "path to file a PEM-encoded SSL private key" });
diff --git a/examples/server/README.md b/examples/server/README.md
index 33a2b95cc..de83ee7d0 100644
--- a/examples/server/README.md
+++ b/examples/server/README.md
@@ -247,7 +247,7 @@ server:
          --host HOST              ip address to listen (default: 127.0.0.1)
          --port PORT              port to listen (default: 8080)
          --path PATH              path to serve static files from (default: )
-         --embedding(s)           enable embedding endpoint (default: disabled)
+         --embedding(s)           restrict to only support embedding use case; use only with dedicated embedding models (default: disabled)
          --api-key KEY            API key to use for authentication (default: none)
          --api-key-file FNAME     path to file containing API keys (default: none)
          --ssl-key-file FNAME     path to file a PEM-encoded SSL private key
diff --git a/gguf-py/gguf/gguf_writer.py b/gguf-py/gguf/gguf_writer.py
index ba6f53cda..2e0b335ee 100644
--- a/gguf-py/gguf/gguf_writer.py
+++ b/gguf-py/gguf/gguf_writer.py
@@ -312,6 +312,8 @@ class GGUFWriter:
         self.add_key_value(key, val, GGUFValueType.STRING)
 
     def add_array(self, key: str, val: Sequence[Any]) -> None:
+        if len(val) == 0:
+            return
         self.add_key_value(key, val, GGUFValueType.ARRAY)
 
     @staticmethod
@@ -845,7 +847,14 @@ class GGUFWriter:
             encoded_val = val.encode("utf-8") if isinstance(val, str) else val
             kv_data += self._pack("Q", len(encoded_val))
             kv_data += encoded_val
-        elif vtype == GGUFValueType.ARRAY and isinstance(val, Sequence) and val:
+        elif vtype == GGUFValueType.ARRAY:
+
+            if not isinstance(val, Sequence):
+                raise ValueError("Invalid GGUF metadata array, expecting sequence")
+
+            if len(val) == 0:
+                raise ValueError("Invalid GGUF metadata array. Empty array")
+
             if isinstance(val, bytes):
                 ltype = GGUFValueType.UINT8
             else:
diff --git a/src/llama.cpp b/src/llama.cpp
index f610a54dc..a4f628905 100644
--- a/src/llama.cpp
+++ b/src/llama.cpp
@@ -4988,6 +4988,7 @@ static void llm_load_hparams(
                 hparams.attn_soft_cap = true;
 
                 switch (hparams.n_layer) {
+                    case 26: model.type = e_model::MODEL_2B; break;
                     case 42: model.type = e_model::MODEL_9B; break;
                     case 46: model.type = e_model::MODEL_27B; break;
                     default: model.type = e_model::MODEL_UNKNOWN;
@@ -11800,6 +11801,7 @@ struct llm_build_context {
 
                 // ref: https://github.com/google/gemma_pytorch/commit/03e657582d17cb5a8617ebf333c1c16f3694670e
                 switch (model.type) {
+                    case e_model::MODEL_2B:
                     case e_model::MODEL_9B:  Qcur = ggml_scale(ctx0, Qcur, 1.0f / sqrtf(float(n_embd_head_k)));   break;
                     case e_model::MODEL_27B: Qcur = ggml_scale(ctx0, Qcur, 1.0f / sqrtf(float(n_embd / n_head))); break;
                     default: GGML_ABORT("fatal error");