recommenders-team/recommenders

[ASK] Abnormal performance when training NPA algorithm

Open

#1260 aperta il 17 dic 2020

Vedi su GitHub
 (1 commento) (0 reazioni) (0 assegnatari)Python (2972 fork)batch import
help wanted

Metriche repository

Star
 (17.706 star)
Metriche merge PR
 (Merge medio 6g 16h) (10 PR mergiate in 30 g)

Descrizione

Description

Hello, an amazing work! Thanks for opensouring all code!

However, when i use NPA algorithm training on a self-made dataset, i find its abnormal performance.

That is, the training loss is decresing, and surprisely the group_auc metric is also decreasing.

The network's performance(group_auc) decrease even at the 2nd epoch, so i think this case is hard to be viewd as overfitting.

Train Settings:

epoch: we have tried 2, 5, 10 word_emb_dim: 1024, here we use the BERT pretrained model to do word embedding, so we adjusted this arg. other args: keep consistant with official setting

initial training loss: 1.6 initial group_auc: 0.54

Other Comments

i also have another problem. What is the meaning of the first evaulation result before training?

Look forward your reply, thanks~

Guida contributor