Update app.py

app.py

@@ -153,7 +153,7 @@ model_d = AutoModelForCausalLM.from_pretrained(
     trust_remote_code=True
 ).eval()
 
-MODEL_ID_P = "strangervisionhf/paddle" #-> https://huggingface.co/PaddlePaddle/PaddleOCR-VL
+MODEL_ID_P = "strangervisionhf/paddle"
 processor_p = AutoProcessor.from_pretrained(MODEL_ID_P, trust_remote_code=True)
 model_p = AutoModelForCausalLM.from_pretrained(
     MODEL_ID_P,
@@ -186,12 +186,21 @@ def generate_image(model_name: str, text: str, image: Image.Image,
 
     images = [image.convert("RGB")]
 
-    messages = [
-        {
-            "role": "user",
-            "content": [{"type": "image"}] + [{"type": "text", "text": text}]
-        }
-    ]
+    # --- ERROR FIX ---
+    # PaddleOCR's processor expects a different message format than the others.
+    # Its chat template expects the 'content' to be a simple string, not a list.
+    if model_name == "PaddleOCR":
+        messages = [
+            {"role": "user", "content": text}
+        ]
+    else:
+        messages = [
+            {
+                "role": "user",
+                "content": [{"type": "image"}] + [{"type": "text", "text": text}]
+            }
+        ]
+        
     prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
     inputs = processor(text=prompt, images=images, return_tensors="pt").to(device)