Refactor leaderboard functionality: update model display to clickable links, improve search logic, and remove unused retrieval_results.json

app.py

@@ -40,25 +40,23 @@ CITATION_BUTTON_TEXT = """
 }
 """
 
-df = load_retrieval_results()
-df[["Model"]] = df[["Model"]].applymap(lambda x: f'<a href="https://huggingface.co/{x}" target="_blank">{x}</a>')
-print(df)
+df = None
 
 def search_leaderboard(model_name):
-    if not model_name:
+    if len(model_name.strip()) == 0:
         return df
     
     threshold = 95  # You can adjust this value to make the search more or less strict
 
     def calculate_similarity(row):
-        similarity = fuzz.partial_ratio(model_name.lower(), row['model'].lower())
+        similarity = fuzz.partial_ratio(model_name.lower(), row["Model"].lower())
         return similarity if similarity >= threshold else 0
 
     # Add a new column for similarity scores
-    df['similarity'] = df.apply(calculate_similarity, axis=1)
+    df["similarity"] = df.apply(calculate_similarity, axis=1)
 
     # Filter and sort the dataframe
-    filtered_df = df[df['similarity'] > 0].sort_values('similarity', ascending=False)
+    filtered_df = df[df["similarity"] > 0].sort_values('similarity', ascending=False)
 
     # Remove the similarity column before returning
     filtered_df = filtered_df.drop('similarity', axis=1)
@@ -67,7 +65,12 @@ def search_leaderboard(model_name):
 
 
 def main():
-    columns_to_show = ["Model", "Model Size", "Embedding Dimension", "Max Tokens", "Web Search Dataset (MRR)", "Web Search Dataset (nDCG@k=None)"]
+    global df
+    df = load_retrieval_results()
+    df[["Model"]] = df[["Model"]].applymap(lambda x: f'<a href="https://huggingface.co/{x}" target="_blank">{x}</a>')
+    df.drop(columns=["Revision", "Precision", "Task"], inplace=True)
+
+    columns_to_show = ["Model", "Model Size", "Embedding Dimension", "Max Tokens", "Num Likes", "Web Search Dataset (Overall Score)"]
     with gr.Blocks() as demo:
         gr.HTML(HEADER)
         
@@ -79,19 +82,30 @@ def main():
                             search_box_retrieval = gr.Textbox(
                                 placeholder="Search for models...", 
                                 label="Search", 
-                                interactive=True
+                                scale=5
+                            )
+                            columns_to_show_input = gr.CheckboxGroup(
+                                label="Columns to Show",
+                                choices=df.columns.tolist(),
+                                value=columns_to_show,
+                                scale=4
                             )
 
                         retrieval_leaderboard = gr.Dataframe(
-                            df[columns_to_show],
+                            value=df[columns_to_show],
                             datatype="markdown",
                             interactive=False
                         )
                         
                         # Submit the search box and the leaderboard
-                        search_box_retrieval.submit(
+                        search_box_retrieval.input(
                             search_leaderboard,
-                            inputs=search_box_retrieval,
+                            inputs=[search_box_retrieval],
+                            outputs=retrieval_leaderboard
+                        )
+                        columns_to_show_input.select(
+                            lambda columns: gr.update(value=df[[col for col in df.columns if col in columns]]),
+                            inputs=columns_to_show_input,
                             outputs=retrieval_leaderboard
                         )
                         

results/retrieval_results.json

@@ -1,18 +0,0 @@
-[
-    {
-        "Model": "intfloat/multilingual-e5-large-instruct",
-        "Revision": "main",
-        "Precision": "f16",
-        "Task": "Retriever",
-        "Num Downloads": 365122,
-        "Num Likes": 279,
-        "License": "mit",
-        "Model Size": 560,
-        "Embedding Dimension": 1024,
-        "Max Tokens": 512,
-        "Web Search Dataset (MRR)": 0.6526209043734583,
-        "Web Search Dataset (nDCG@k=None)": 0.7165619055124024,
-        "Web Search Dataset (Recall@5)": 0.7414125820146662,
-        "Web Search Dataset (Overall Score)": 0.7035317973001757
-    }
-]