diff --git a/example-cpu.py b/example-cpu.py
index a71e718..7311308 100644
--- a/example-cpu.py
+++ b/example-cpu.py
@@ -11,6 +11,7 @@ import json
 from pathlib import Path
 from llama import ModelArgs, Transformer, Tokenizer, LLaMA
 
+
 def load(
         ckpt_dir: str,
         tokenizer_path: str,
@@ -42,6 +43,7 @@ def load(
     print(f"Loaded models in {time.time() - start_time:.2f} seconds")
     return generator
 
+
 def main(
         ckpt_dir: str = './model',
         tokenizer_path: str = './tokenizer/tokenizer.model',
@@ -94,5 +96,6 @@ def main(
 
     print(f"Inference took {time.time() - start_time:.2f} seconds")
 
+
 if __name__ == "__main__":
     fire.Fire(main)