diff --git a/llama.cpp b/llama.cpp
index fb71aef19..e8522dbe8 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -995,6 +995,11 @@ static void llama_model_load_internal(
     std::unique_ptr<llama_model_loader> ml(new llama_model_loader(fname, use_mmap));
 
     vocab = std::move(ml->file_loader->vocab);
+
+    if (vocab_only) {
+        return;
+    }
+
     model.hparams = ml->file_loader->hparams;
     model.n_gpu_layers = n_gpu_layers;
     llama_file_version file_version = ml->file_loader->file_version;
@@ -1055,10 +1060,6 @@ static void llama_model_load_internal(
         }
     }
 
-    if (vocab_only) {
-        return;
-    }
-
     auto & ctx = model.ctx;
 
     size_t ctx_size;
diff --git a/models/ggml-vocab.bin b/models/ggml-vocab.bin
index 38f63493a..c1a3194c4 100644
Binary files a/models/ggml-vocab.bin and b/models/ggml-vocab.bin differ