LengSicong/Tell2Design

Wish to provide more data thanks to.

ZijinQiu opened this issue · 3 comments

Noticed in the paper that there are only 4k manually labelled data, aren't there more than 5k, easily available, thanks!

Please refer to this issue

78351edf5f3a6b7020c6ef2f7c57f43
Thank you for your reply. When reproducing the code, we found that there are data such as near_x_min, etc. These four data are not explained in the paper. What do they mean? I hope to get your help~

And what model is used to generate the intelligent annotations in the article? T5?