Open chris-opendata opened 1 year ago
Thanks for your interest in our work and your question. Due to the maximum length limitation of the model deployed by GPTScore, in-context learning is challenging to apply in the text summarization task. We did not construct demonstration samples for the summarization task as we did for other tasks.
Great work. I intend to utilize it for my abstractive text summarization paper. Would you be able to upload the related examples? I am interested in evaluating semantic coverage, factuality, informativeness, and fluency following the aspect definitions in your paper.
Kind regards