search

HsuWanTing / unified-summarization

Official codes for the paper: A Unified Model for Extractive and Abstractive Summarization using Inconsistency Loss.
https://hsuwanting.github.io/unified_summ/
MIT License
125 stars 30 forks source link

Difference between SummaRuNNer and extractor of unified-summarization #14

Closed bckim92 closed 5 years ago

bckim92 commented 5 years ago

Hi, thanks for your great work.

I get to know that there is an architecture difference between SummaRuNNer and extractor of this work. In SummaRuNNer, they add absolute/relative positional embeddings to the classification layer, but in your work, you didn't add those embeddings (related code). Is there any special reason for this architecture choice (e.g. faster training, performance degradation)? If so, can you tell it to me?

Thanks!

HsuWanTing commented 5 years ago

Hi, Actually, we have tried adding the absolute and relative positional embeddings in our model. But we found those embeddings doesn’t help (accuracy doesn’t improve). So we just deleted them for simplicity.

bckim92 commented 5 years ago

Okay, thanks for your reply!