2025-12-25 07:55:08,518 - INFO - Loaded dataset: math-500
2025-12-25 07:55:08,527 - INFO - Loaded 100 profiles from ../data/complex_profiles_v2/profiles_100.jsonl
2025-12-25 07:55:08,528 - INFO - Running method: rag_vector
2025-12-25 07:55:08,528 - INFO -   Profile 1/2

Generating test split:   0%|          | 0/500 [00:00<?, ? examples/s]
Generating test split: 100%|██████████| 500/500 [00:00<00:00, 9910.65 examples/s]
2025-12-25 07:55:12,047 - ERROR - Error in session: No module named 'json_repair'
2025-12-25 07:55:12,050 - ERROR - Error in session: No module named 'json_repair'
2025-12-25 07:55:12,050 - INFO -   Profile 2/2
2025-12-25 07:55:12,053 - ERROR - Error in session: No module named 'json_repair'
2025-12-25 07:55:12,056 - ERROR - Error in session: No module named 'json_repair'
Traceback (most recent call last):
  File "/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/collaborativeagents/scripts/run_experiments.py", line 488, in <module>
    main()
  File "/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/collaborativeagents/scripts/run_experiments.py", line 476, in main
    analysis = runner.run_all()
               ^^^^^^^^^^^^^^^^
  File "/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/collaborativeagents/scripts/run_experiments.py", line 292, in run_all
    analysis = self._analyze_results(all_results)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/projects/bfqt/users/yurenh2/ml-projects/personalization-user-model/collaborativeagents/scripts/run_experiments.py", line 365, in _analyze_results
    best = max(values, key=values.get)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: max() arg is an empty sequence