Add fallback sorting when no task results exist

src/populate.py

@@ -16,8 +16,15 @@ def get_leaderboard_df(results_path: str, requests_path: str, cols: list, benchm
 
     df = pd.DataFrame.from_records(all_data_json)
     # Sort by the first task (EMEA NER) since we don't have an average for NER tasks
-    first_task = list(Tasks)[0]  # emea_ner
-    df = df.sort_values(by=[getattr(AutoEvalColumn, first_task.name).name], ascending=False)
+    # If no results exist yet, just sort by model name
+    if not df.empty:
+        first_task = list(Tasks)[0]  # emea_ner
+        task_col_name = getattr(AutoEvalColumn, first_task.name).name
+        if task_col_name in df.columns:
+            df = df.sort_values(by=[task_col_name], ascending=False)
+        else:
+            # Fallback to sorting by model name if no task results yet
+            df = df.sort_values(by=[AutoEvalColumn.model.name], ascending=True)
     df = df[cols].round(decimals=2)
 
     # filter out if any of the benchmarks have not been produced