file path

app.py

@@ -1,11 +1,10 @@
-import re
+import os
 import gradio as gr
 import numpy as np
-from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
+from transformers import Qwen2VLForConditionalGeneration, AutoProcessor
 from qwen_vl_utils import process_vision_info
 import torch
 from ast import literal_eval
-from PIL import Image
 
 # Load the model on the available device(s)
 model = Qwen2VLForConditionalGeneration.from_pretrained(
@@ -42,17 +41,14 @@ tax_deductions = '''Extract the following information in the given format:
                 }
 '''
 
-def demo(image, prompt):
-    # Convert image to bytes or suitable format if needed
-    image_name = image  # If you're directly using the image
-
+def demo(image_path, prompt):
     messages = [
         {
             "role": "user",
             "content": [
                 {
                     "type": "image",
-                    "image": image_name,
+                    "image": image_path,  # Use the file path here
                 },
                 {"type": "text", "text":  prompt},
             ],
@@ -91,9 +87,12 @@ def demo(image, prompt):
     return json
 
 def process_document(image):
-    # Pass the image directly
-    one = demo(image, other_benifits)
-    two = demo(image, tax_deductions)
+    # Save the uploaded image temporarily and get its path
+    image_path = image.name  # Gradio provides an interface to access the file name
+
+    # Process the image with your model
+    one = demo(image_path, other_benifits)
+    two = demo(image_path, tax_deductions)
     json_op = {
         "tax_deductions": one,
         "other_benifits": two