fix: do not rank but strike out methods that cannot run for certain subsets.

main.py

@@ -206,7 +206,7 @@ def make_overview_leaderboard(lbs: [LBContainer]):
             missing_models = all_models.difference(df_models)
             if missing_models:
                 missing_models_df = pd.DataFrame(
-                    [[mm, len(df_models) + 1] for mm in missing_models],
+                    [[mm, "--"] for mm in missing_models],
                     columns=["Model", lb.name],
                 )
                 df = pd.concat([df, missing_models_df], ignore_index=True)
@@ -220,7 +220,7 @@ def make_overview_leaderboard(lbs: [LBContainer]):
     def highlight_top3(col):
         styles = [""] * len(col)
         for index_i in range(len(col)):
-            if col.iloc[index_i] <= 3:
+            if (not isinstance(col.iloc[index_i], str)) and col.iloc[index_i] <= 3:
                 styles[index_i] = (
                     f"background-color: {medal_colors[col.iloc[index_i] - 1]};"
                 )