Update app.py

app.py

@@ -391,6 +391,10 @@ def format_sparse_vector_output(splade_vector, tokenizer, is_binary=False):
     else:
         terms_list = []
         for i, (term, weight) in enumerate(sorted_representation):
+            # Limiting to 50 terms for display to avoid overly long output
+            if i >= 50:
+                terms_list.append(f"...and {len(sorted_representation) - 50} more terms.")
+                break
             if is_binary:
                 terms_list.append(f"**{term}**")
             else:
@@ -552,31 +556,47 @@ with gr.Blocks(title="SPLADE Demos", css=css) as demo:
                 "Binary Bag-of-Words"
             ]
 
+            # Input components for the second tab
+            query_model_radio = gr.Radio(
+                model_choices,
+                label="Choose Query Encoding Model",
+                value="MLM encoder (SPLADE-cocondenser-distil)"
+            )
+            doc_model_radio = gr.Radio(
+                model_choices,
+                label="Choose Document Encoding Model",
+                value="MLM encoder (SPLADE-cocondenser-distil)"
+            )
+            query_text_input = gr.Textbox(
+                lines=3,
+                label="Enter Query Text:",
+                placeholder="e.g., famous dishes of Padua"
+            )
+            doc_text_input = gr.Textbox(
+                lines=5,
+                label="Enter Document Text:",
+                placeholder="e.g., Padua's cuisine is as famous as its legendary University."
+            )
+
+            # --- MODIFIED: Output component as a Textbox with scrollbar ---
+            output_dot_product_text = gr.Textbox(
+                label="Dot Product Result and Representations",
+                lines=20,  # This controls the initial height
+                autoscroll=True, # Automatically scroll to bottom if new content is added
+                interactive=False, # Make it non-editable
+                max_lines=None # Allows infinite scrolling (no max height)
+            )
+
+            # Update the gr.Interface call to use the new Textbox output
             gr.Interface(
                 fn=calculate_dot_product_and_representations_independent,
                 inputs=[
-                    gr.Radio(
-                        model_choices,
-                        label="Choose Query Encoding Model",
-                        value="MLM encoder (SPLADE-cocondenser-distil)"
-                    ),
-                    gr.Radio(
-                        model_choices,
-                        label="Choose Document Encoding Model",
-                        value="MLM encoder (SPLADE-cocondenser-distil)"
-                    ),
-                    gr.Textbox(
-                        lines=3,
-                        label="Enter Query Text:",
-                        placeholder="e.g., famous dishes of Padua"
-                    ),
-                    gr.Textbox(
-                        lines=5,
-                        label="Enter Document Text:",
-                        placeholder="e.g., Padua's cuisine is as famous as its legendary University."
-                    )
+                    query_model_radio,
+                    doc_model_radio,
+                    query_text_input,
+                    doc_text_input
                 ],
-                outputs=gr.Markdown(),
+                outputs=output_dot_product_text, # Changed to the Textbox
                 allow_flagging="never"
             )