Changed some values to be updated

app.py

@@ -15,47 +15,40 @@ model.to(device)
 labels = model.config.id2label
 print(labels)
 
-# helper function to unnormalize bounding boxes 
-def unnormalize_box(bbox, width, height): 
-    return [
-        width * (bbox[0] / 1000), 
-        height * (bbox[1] / 1000), 
-        width * (bbox[2] / 1000), 
-        height * (bbox[3] / 1000),
-    ]
-
 # process the image in the correct format
 # extract token classifications 
 def parse_ticket_image(image): 
+
+    # Process image
     if image: 
         document = image.convert("RGB") if image.mode != "RGB" else image
     else: 
         print(f"Warning - no image or malformed image!") 
         return pd.DataFrame()
 
+    # Encode document image
     encoding = processor(document, return_tensors="pt", truncation=True)
 
+    # Move encoding to appropriate device
     for k, v in encoding.items(): 
         encoding[k] = v.to(device)
 
+    # Perform inference
     outputs = model(**encoding)
 
+    # extract predictions
     predictions = outputs.logits.argmax(-1).squeeze().tolist()
-    token_boxes = encoding.bbox.squeeze().tolist()
 
     input_ids = encoding.input_ids.squeeze().tolist()
     words = [processor.tokenizer.decode(id) for id in input_ids]
 
-    width, height = document.size
-    true_predictions = []
-    true_boxes = []
+    extracted_fields = []
 
     for idx, pred in enumerate(predictions): 
         label = model.config.id2label[pred]
         # apparently 'O' stands for non-entity tokens
         if label != 'O': 
-            true_predictions.append(label)
-            true_boxes.append(unnormalize_box(token_boxes[idx], width, height))
+            extracted_fields.append((label, words[idx]))
 
     if len(extracted_fields) == 0:
         print(f"Warning - no fields were extracted!") 
@@ -67,8 +60,8 @@ def parse_ticket_image(image):
     values = values[:min_length]
     
     data = {
-        "Field": true_predictions,
-        "Value": words[1:len(true_predictions)+1]
+        "Field": fields,
+        "Value": values
     }
     df = pd.DataFrame(data)