Add debug at line 741

evaluation/evaluate_all_pai.py

@@ -737,6 +737,8 @@ def run_evaluation(output_file, tasks=None, grouping="per-dataset", silent_eval=
 
     print(f"[DEBUG_AFTER_ANALYZE] Got {len(qa_type_counts)} QA types", flush=True)
     sys.stdout.flush()
+    print(f"[DEBUG] Line 741 - About to determine tasks, tasks={tasks}", flush=True)
+    sys.stdout.flush()
 
     # Determine which tasks to run
     if tasks is None: