Skip to content

evaluate question #42

@d1jiasheng

Description

@d1jiasheng

您好,我在跑豆瓣的data.pkl数据集的时候,在save_step会卡住,我初步分析原因应该是验证集数量太大和for循环太多导致的。我观察了在跑data_small.pkl的时候,验证集的batch_num大概是31个,验证需要花费时间大概两分钟,而data.pkl的batch_num大概是1562个,验证的时候就会一直卡在那个地方。我想问一下您在实验的时候对这个部分有什么优化吗

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