Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Language Understanding and Reasoning on MMLU, PIQA, HellaSwag, WinoGrande, ARC-Challenge

46.45MMLU (5s)

BF16

22.186828.485934.78541.0841Feb 16, 2024
Updated 4d ago

Evaluation Results

MethodLinks
2024.02
46.4577.8657.1468.3543.3458.63
2024.02
43.6576.9955.3868.3541.2157.12
2024.02
41.3277.2654.7468.3540.6156.46
2024.02
41.2277.4754.4167.0939.0855.85
2024.02
39.6876.2755.1467.5640.6155.85
2024.02
39.5775.4651.6867.1638.3954.45
2024.02
38.6575.2453.767.3238.5654.69
2024.02
29.2573.6148.761.0933.2749.18
2024.02
25.4258.9229.250.8319.4536.76
2024.02
25.3852.3925.750.1221.3334.98
2024.02
24.1253.4326.3349.9621.5835.08
2024.02
23.6270.0843.7961.6429.0945.64
2024.02
23.1249.5125.0449.5722.6933.99