OpenEval

该数据集旨在对代码生成模型进行更广泛的评估，与HumanEval相似。它主要针对的任务是代码生成。该数据集不仅提供了丰富的测试案例，还旨在评估模型在生成代码时的多样性和创造性。通过对这一数据集的深入研究和分析，研究人员和开发者可以更好地理解代码生成模型的性能和局限性。

This dataset aims to conduct a broader evaluation of code generation models, similar to HumanEval. Its primary targeted task is code generation. This dataset not only provides abundant test cases, but also aims to evaluate the diversity and creativity of models during code generation. Through in-depth research and analysis on this dataset, researchers and developers can gain a better understanding of the performance and limitations of code generation models.