Update app.py

app.py

@@ -14,10 +14,8 @@ from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, Auto
 def save_temp_image(image: Image.Image) -> str:
     """
     Saves the given PIL Image object as a temporary PNG file.
-
     Args:
         image (Image.Image): The image to be saved.
-
     Returns:
         str: The file path of the saved temporary image.
     """
@@ -34,13 +32,13 @@ def id_extractor(image: Image.Image) -> Dict:
 
     model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
 
-        "Qwen/Qwen2.5-VL-3B-Instruct", torch_dtype="auto", device_map="auto"
+        "Qwen/Qwen2.5-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
 
     )
 
     # default processer
 
-    processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-3B-Instruct")
+    processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct")
 
     messages = [
 
@@ -106,8 +104,9 @@ def id_extractor(image: Image.Image) -> Dict:
         generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
 
     )
-    return output_text
-
+    resp = output_text[-1].replace("```json", "").replace("```", "").strip()
+    return json.loads(resp)
+ 
 # Define the Gradio interface for the ID extractor
 id_interface = gr.Interface(
     fn=id_extractor,
@@ -118,4 +117,4 @@ id_interface = gr.Interface(
 )
 
 # Launch the Gradio interface
-id_interface.launch(mcp_server=True)
+id_interface.launch(mcp_server=True)