thr

app.py

@@ -38,6 +38,7 @@ def model_inference(
     # print(images)
     # print(type(images))
     images = [{"type": "image", "image": Image.open(image[0])} for image in images]
+    images.append({"type": "text", "text": text})
     print(images)
     # model = Qwen2VLForConditionalGeneration.from_pretrained(
     # "Qwen/Qwen2-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
@@ -57,7 +58,7 @@ def model_inference(
     messages = [
         {
             "role": "user",
-            "content": images.append({"type": "text", "text": text}),
+            "content": images,
         }
     ]