Skip to content

Discrepancy in eval_end2end_ace2005e #18

@InfiniH

Description

@InfiniH

Hi,
I'm encountering a puzzling problem while running the eval_end2end_ace2005e.py with the DEGREE_e2e_ace05e.mdl pretrained model. The evaluation results I'm obtaining are significantly different from the results reported in the associated paper, even though I have not made any modifications to the model itself.
Has anyone else encountered a similar problem when running the eval_end2end_ace2005e with pretrained model?
Could you kindly provide insights into potential areas that might lead to such discrepancies? Are there any data processing intricacies that could be causing this?
Any assistance in troubleshooting this matter would be greatly appreciated.
image

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