BERT-whitening-pytorch icon indicating copy to clipboard operation
BERT-whitening-pytorch copied to clipboard

Pytorch version of BERT-whitening

Results 12 BERT-whitening-pytorch issues
Sort by recently updated
recently updated
newest added

大佬,在[all_utils.py](github.com/autoliuweijie/BERT-whitening-pytorch/blob/main/all_utils.py)的函数[compute_kernel_bias](github.com/autoliuweijie/BERT-whitening-pytorch/blob/b5cfbd606bd19fc3b3adf9e074dc0bfd830ef597/all_utils.py#L63),是要将语料们先通过BERT池化成句向量后拼接成vecs来输入。 假设一个场景是给定查询的语句q,也给定待匹配的语句库,是不是每次都要将q和语句库中所有语句一起做上述操作才行?如果是,这似乎计算量好大。

Hi, I downloaded the datasets as you described, but the directory /data/downstream is different from what it should be. This is what I geted: ![2](https://user-images.githubusercontent.com/51829876/205433652-ed7543ea-93a7-4f34-b035-713e3b07e0b9.jpg)