HumanEval+
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HumanEval+ (test)
98.1Pass@1
132
HumanEval+ v1 (test)
87.8Pass Rate
55
HumanEval+ Out-of-Domain (test)
81.71Accuracy
18
HumanEval+
97Pass@16
15
HumanEval+
34.76Score
11
HumanEval+ (test)
1.27Error Rate
7
HumanEval+ ko
92.1Score
3