Update app.py

app.py

@@ -1,40 +1,36 @@
-import numpy as np
-import pandas as pd
-import keras
-import gradio as gr
-
-
-
-import os
-os.environ["KERAS_BACKEND"] = "tensorflow"
-	
-import keras
-
-model = keras.saving.load_model("hf://kim1688/casting_defect_resnet50")
-
-
-
-
-def upload_images(image_paths):
-  df = pd.DataFrame(columns=["Index", "File", "Result"])
-  for i in range(len(image_paths)):
-    df.loc[i] = [str(i+1), image_paths[i].split("\\")[-1], predict(image_paths[i])]
-  return df
-
-# Function to preprocess image and predict
-def predict(image_path):
-  img = keras.utils.load_img(image_path, target_size=(300, 300))
-  img_array = keras.utils.img_to_array(img)
-  img_array = keras.ops.expand_dims(img_array, 0)
-  prediction = model.predict(img_array)
-  class_names = ["Defective", "Normal"]  # Class 0: def, Class 1: ok
-  predicted_class = class_names[1] if prediction > 0.5 else class_names[0]
-  return predicted_class
-
-demo = gr.Interface(
-    upload_images, 
-    gr.File(file_count="multiple", file_types=[".jpg", ".jpeg", ".png", ".bmp", ".tif", ".tiff"]), 
-    gr.DataFrame(headers=["Index", "File", "Result"])
-)
-
+import os
+import numpy as np
+import pandas as pd
+import keras
+import gradio as gr
+
+os.environ["KERAS_BACKEND"] = "tensorflow"
+print("loading file")
+model = keras.saving.load_model("hf://kim1688/casting_defect_resnet50")
+
+
+
+
+def upload_images(image_paths):
+  df = pd.DataFrame(columns=["Index", "File", "Result"])
+  for i in range(len(image_paths)):
+    df.loc[i] = [str(i+1), image_paths[i].split("\\")[-1], predict(image_paths[i])]
+  return df
+
+# Function to preprocess image and predict
+def predict(image_path):
+  img = keras.utils.load_img(image_path, target_size=(300, 300))
+  img_array = keras.utils.img_to_array(img)
+  img_array = keras.ops.expand_dims(img_array, 0)
+  prediction = model.predict(img_array)
+  class_names = ["Defective", "Normal"]  # Class 0: def, Class 1: ok
+  predicted_class = class_names[1] if prediction > 0.5 else class_names[0]
+  return predicted_class
+
+demo = gr.Interface(
+    upload_images, 
+    gr.File(file_count="multiple", file_types=[".jpg", ".jpeg", ".png", ".bmp", ".tif", ".tiff"]), 
+    gr.DataFrame(headers=["Index", "File", "Result"])
+)
+
 demo.launch(share=True, debug=True)