m2e2 evaluation

[bernardoacalvo]

Hi,

Thank you for sharing your work.
However, I think the evaluation of the m2e2 dataset is not totally clear.

In your paper, in table 3, there are the evaluation results.

When considering only the multimedia training row, therefore those models have as input a multimodal document, what does "Text-Only Evaluation" mean?
Is it evaluating on the "text_only_event.json" file?
Is it evaluating in both "text_only_event.json" and "text_multimedia_event.json" files?

Thanks.

[seebergerph]

Hi, were you able to reproduce the results? Unfortunately, I got very low precision scores for text argument roles when I ran the code as it is.

[York-Gold]

我也有同样的问题，谢谢！