Code change : Load model before prediction for faster inference

app.py

@@ -6,18 +6,17 @@ import torch
 
 from transformers import DonutProcessor, VisionEncoderDecoderModel
 
-def predict(inp):
-
-    # Check GPU
-    device = "cuda" if torch.cuda.is_available() else "cpu"
+# Check GPU
+device = "cuda" if torch.cuda.is_available() else "cpu"
     
-    # Load processor
-    processor = DonutProcessor.from_pretrained("jonathanjordan21/donut_fine_tuning_food_composition_id")
+# Load processor
+processor = DonutProcessor.from_pretrained("jonathanjordan21/donut_fine_tuning_food_composition_id")
     
-    # Load model
-    model = VisionEncoderDecoderModel.from_pretrained("jonathanjordan21/donut_fine_tuning_food_composition_id")
+# Load model
+model = VisionEncoderDecoderModel.from_pretrained("jonathanjordan21/donut_fine_tuning_food_composition_id")
 
 
+def predict(inp):
     # Define Json Parser
     def get_komposisi(image_path, image=None):
         image = Image.open(image_path).convert('RGB') if image== None else image.convert('RGB')
@@ -50,9 +49,6 @@ def predict(inp):
     return out
 
 
-def upload_file(files):
-    file_paths = [file.name for file in files]
-    return file_paths
 
 gr.Interface(fn=predict,
              inputs=gr.Image(type="pil"),