bug: set eval mode during inference

main.py

@@ -1,27 +1,22 @@
 #!/usr/bin/env python3
 
+import argparse
 import io
+import multiprocessing
 import os
 import time
-import argparse
 from distutils.util import strtobool
 from typing import Union
+
 import cv2
 import numpy as np
 import torch
-
 from flask import Flask, request, send_file
 from flask_cors import CORS
-from lama_cleaner.helper import (
-    download_model,
-    load_img,
-    norm_img,
-    resize_max_size,
-    numpy_to_bytes,
-    pad_img_to_modulo,
-)
 
-import multiprocessing
+from lama_cleaner.helper import (download_model, load_img, norm_img,
+                                 numpy_to_bytes, pad_img_to_modulo,
+                                 resize_max_size)
 
 NUM_THREADS = str(multiprocessing.cpu_count())
 
@@ -102,6 +97,7 @@ def run(image, mask):
     mask = torch.from_numpy(mask).unsqueeze(0).to(device)
 
     start = time.time()
+    with torch.no_grad():
         inpainted_image = model(image, mask)
 
     print(f"process time: {(time.time() - start)*1000}ms")
@@ -135,6 +131,7 @@ def main():
 
     model = torch.jit.load(model_path, map_location="cpu")
     model = model.to(device)
+    model.eval()
     app.run(host="0.0.0.0", port=args.port, debug=args.debug)