Skip to content

question about dataset #14

@XpastaX

Description

@XpastaX

Thank you for your excellent work! I have some questions regarding the dataset.

Is there a detailed introduction section for the dataset? What does each key in a sample dict represent? Specifically, could you clarify what the target code and the fastest code in the test data represent? For example, is the target code the best solution submitted by the user who also provided the source code, or is it the best solution submitted by any user for this problem? Similarly, how is the fastest code defined?

Additionally, do I need to rerun all evaluation results on the test set, including the source code and target code? I'm concerned that the CPU time of the code might differ from your reported results, even with the provided Docker environment, as evaluations may run on different devices.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions