huawei-noah/vega

searching valid metric is better than re-train valid metric in autoFIS?

オープン

#107 opened on 2021/04/08

 (2 件のコメント) (0 件のリアクション) (1 人の担当者)Python (177 件のフォーク)auto 404
good first issuequestion

Repository metrics

Stars
 (848 個のスター)
PR merge metrics
 (30d に merged PR はありません)

説明

Thank you for your excellent codes. I run your autoFIS codes (autoDeepFM), but found that the valid metric after searching (auc=0.7896248002165958) is better than the retrain valid metric (auc=0.7851094769386633) in the log. i.e. [searching valid metric] > [re-train valid metric] > [official valid metric], it looks like the performance improvement is achieved by BN but autofis damages the valid metric. Something wrong ?

the valid metric after searching: image the valid metric after re-train: image

コントリビューターガイド