Update app.py

app.py

@@ -6,8 +6,8 @@ from transformers import MllamaForConditionalGeneration, AutoProcessor
 from PyPDF2 import PdfReader
 import tempfile
 import os
-
-token=os.getenv("HF_TOKEN")
+from pdf2image import convert_from_path
+token= os.getenv("HF_TOKEN")
 # Model and processor setup
 model_id = "meta-llama/Llama-3.2-11B-Vision-Instruct"
 
@@ -22,19 +22,11 @@ processor = AutoProcessor.from_pretrained(model_id)
 
 def process_pdf(pdf_file):
     """Extract text from each page of a PDF."""
-    # Save the uploaded PDF to a temporary file
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_pdf:
-        temp_pdf.write(pdf_file.read())
-        temp_pdf_path = temp_pdf.name
-
-    # Read the PDF
-    pdf_reader = PdfReader(temp_pdf_path)
+    # Read the PDF using pdf2image to convert pages to images
+    images = convert_from_path(pdf_file.name)
     extracted_text = {}
 
-    for i, page in enumerate(pdf_reader.pages):
-        # Convert page content to an image (placeholder: replace with actual PDF-to-image conversion)
-        page_image = Image.new("RGB", (1000, 1000), color=(255, 255, 255))
-        
+    for i, page_image in enumerate(images):
         # Define the instruction for OCR
         messages = [
             {"role": "user", "content": [