model loads on request

inference_utils.py

@@ -90,7 +90,6 @@ def get_predictions(model,img,return_raw = False):
                 "confidenceInClass" : float(i[4]),
                 "DetectedClass" : every_monument[int(i[5])]
                 }
-        print(temp)
         results.append(temp)
     
     # changing the final arrays of dictionaries before returning

main.py

@@ -4,15 +4,18 @@ import cv2
 import torch
 import os
 import aiofiles
+from functools import lru_cache
 
+@lru_cache(maxsize=1)
+def load_model():
+    weights_path = os.path.join('best.pt')
+    yolo_path = os.path.join('yolov5')
+    model = torch.hub.load(yolo_path, 'custom', path = weights_path, source = 'local',device='cpu',force_reload=True)
 
-weights_path = os.path.join('best.pt')
-yolo_path = os.path.join('yolov5')
-model = torch.hub.load(yolo_path, 'custom', path = weights_path, source = 'local',device='cpu',force_reload=True)
-
+    return model
+    
 
 app = FastAPI()
-
 CHUNK_SIZE = 1024 * 1024
 
 @app.get("/")
@@ -21,7 +24,7 @@ async def root():
 
 @app.post("/detect-monument/")
 async def upload(file: UploadFile = File(...)):
-    
+    model = load_model()
     try:
         filepath = os.path.join('./', os.path.basename(file.filename))
         async with aiofiles.open(filepath, 'wb') as f: