update

app.py

@@ -14,7 +14,7 @@ model = AutoModelForImageTextToText.from_pretrained(
 ).to("cuda").eval()
 
 @spaces.GPU
-def model_inference(input_dict, history):
+def inference(input_dict, history):
     text = input_dict["text"]
     files = input_dict["files"]
 
@@ -85,15 +85,9 @@ def model_inference(input_dict, history):
         buffer += new_text
         yield buffer
 
-with gr.Blocks() as demo:
-    gr.ChatInterface(
-        fn=model_inference,
-        description="# **VL-Rethinker-7B**",
-        fill_height=True,
-        textbox=gr.MultimodalTextbox(label="Query Input", file_types=["image"], file_count="multiple"),
-        stop_btn="Stop Generation",
+demo = gr.ChatInterface(
+        fn=inference,
         multimodal=True,
-        cache_examples=False,
     )
 
 demo.launch(debug=True)