{ "mmlu": 0.2290, "mmlu_baseline": 0.25, "hellaswag": 0.2668, "hellaswag_baseline": 0.25, "arc_challenge_acc": 0.1843, "arc_challenge_acc_baseline": 0.25, "arc_challenge_acc_norm": 0.2338, "arc_challenge_acc_norm_baseline": 0.25, "boolq": 0.5590, "boolq_baseline": 0.50, "winogrande": 0.5146, "winogrande_baseline": 0.50, "gsm8k_exact_match_5shot": 0.0091, "gsm8k_exact_match_5shot_baseline": 0.00 }