test input

app.py

@@ -108,7 +108,7 @@ def deepseek(image, text_input):
     det_pattern = r"<\|det\|>\[\[(.+)]]<\|\/det\|>"
 
     det_content = re.search(det_pattern, answer).group(1)
-    bbox = [ int(v.trim()) / 999 for v in det_content.split(",")]
+    bbox = [ int(v.strip()) / 999 for v in det_content.split(",")]
     w, h = image.size()
     bbox = [bbox[0] * w, bbox[1] * h, bbox[2] * w, bbox[3] * h] 
     
@@ -181,7 +181,7 @@ css = """
 with gr.Blocks(css=css) as demo:
     gr.Markdown(
     """
-    # Demo for OS-ATLAS: A Foundation Action Model For Generalist GUI Agents
+    # Demo for Deepseek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
     """)
     with gr.Row():
         with gr.Column():