You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have two questions regarding the document generation task, i.e., Table 1 in your paper.
Which size of pre-trained GPT-2 did you use to compute MAUVE? In the original MAUVE paper, they tried different sizes.
It seems to me that the length of the generated output matters when we try to alleviate the repetition problem, i.e., shorter, less repetitive. Could you recall the average length of the generated outputs?
Thank you in advance and looking forward to hearing from you.
Best,
Dong
The text was updated successfully, but these errors were encountered:
Hi there,
Thanks for releasing the codes. Much appreciated.
I have two questions regarding the document generation task, i.e., Table 1 in your paper.
Thank you in advance and looking forward to hearing from you.
Best,
Dong
The text was updated successfully, but these errors were encountered: