GSM8K, GSM-Hard, MATH, StrategyQA, Bamboogle, Coin, Letter, LB, Headline
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
GSM8K, GSM-Hard, MATH, StrategyQA, Bamboogle, Coin, Letter, LB, Headline (test)
92.6GSM8K Accuracy
15
Loading the SOTA2 catalog…