MMLU, GSM8k, HellaSwag, WinoGrande
Benchmarks
TaskDatasetSOTA resultTrendResults
MMLU, GSM8k, HellaSwag, WinoGrande
78.9Average Score
58
MMLU, GSM8k, HellaSwag, WinoGrande
72.98MMLU Accuracy
17
MMLU, GSM8k, HellaSwag, WinoGrande (test)
77.18MMLU Accuracy
13
MMLU, GSM8k, HellaSwag, WinoGrande (test)
86.55MMLU Accuracy
13