Why manually manage L2 norm multiplication of model parameters ?

shenweichen / DeepCTR-Torch

【PyTorch】Easy-to-use,Modular and Extendible package of deep-learning based CTR models.

https://deepctr-torch.readthedocs.io/en/latest/index.html

Apache License 2.0

2.95k stars 696 forks source link

Why manually manage L2 norm multiplication of model parameters ? #279

Open zhijiejia opened 1 year ago

zhijiejia commented 1 year ago

Please refer to the FAQ in doc and search for the related issues before you ask the question.

Describe the question(问题描述) In Class: BaseModel, Why manually manage L2 norm multiplication of model parameters ? Why not use the optimizer to achieve this ? Is it set up like this on purpose?

Operating environment(运行环境):

deepctr-torch version [0.2.9]

zhijiejia commented 1 year ago

Why the dense feature can have gt 1 dimension ? Can provide a example to prove it ?