search

reczoo / RecZoo

A curated model zoo for recommendation tasks
Apache License 2.0
163 stars 38 forks source link

关于Loss的问题 #23

Closed D-JI-Feng closed 1 year ago

D-JI-Feng commented 1 year ago

作者你好，我在论文中看到，loss是这样求的：L = Lo + lamda * Lc + gamma Li，但在代码里，似乎loss是这样求的：L = Lc + gamma_ parameters.normloss() + lamda * Li。有些不一致，可以解释一下吗？

kyriemao commented 1 year ago

Hi,

Please refer to this issue https://github.com/xue-pai/UltraGCN/issues/12.