Benchmark

Instruct HumanEval

general text

Instruction-based variant of HumanEval benchmark for evaluating large language models' code generation capabilities with functional correctness using pass@k metric on programming problems

语言EN
满分1
参评模型1

模型排名

名次 模型 机构 分数 来源
1 Llama 3.1 Nemotron 70B Instruct NVIDIA 73.8 来源 ↗