diff --git a/examples/models/llama2/export_llama_lib.py b/examples/models/llama2/export_llama_lib.py
index 885850988d..268b68cc2e 100644
--- a/examples/models/llama2/export_llama_lib.py
+++ b/examples/models/llama2/export_llama_lib.py
@@ -352,11 +352,11 @@ def _export_llama(modelname, args) -> str:  # noqa: C901
         if (
             args.dtype_override == "fp16" and metadata["get_dtype"] != 5
         ) or args.quantization_mode == "int4":
-            print(f"model.to torch.float16")
+            print("model.to torch.float16")
             model = model.to(dtype=torch.float16)
             metadata["get_dtype"] = 5
         elif args.dtype_override == "fp32" and metadata["get_dtype"] != 6:
-            print(f"model.to torch.float32")
+            print("model.to torch.float32")
             model = model.to(dtype=torch.float32)
             metadata["get_dtype"] = 6
         else: