Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Do Transformer Modifications Transfer Across Implementations and Applications? #2109

Open
icoxfog417 opened this issue Jun 18, 2022 · 1 comment

Comments

@icoxfog417
Copy link
Member

一言でいうと

Transformerに対する変更がどれだけNLPタスクの精度向上に貢献するか調査した研究。活性関数、レイヤ正規化、深さ、エンコーディング(factorizationなど)、レイヤ間パラメーター共有、Softmax、Transformerの構造に対する各種変更について、タスク、ハイパーパラメーター、最適化アルゴリズム、パラメーター数、1秒あたりFLOPを共通化して検証している。結果として、精度の改善がみられるのは活性化関数や正規化などの細かい変更、パラメーター数や計算量を増やした場合であった。ハイパーパラメーターチューニングがあまり効果がなかったとも言及されている。

論文リンク

https://arxiv.org/abs/2102.11972

著者/所属機関

Sharan Narang, Hyung Won Chung, Yi Tay, William Fedus, Thibault Fevry, Michael Matena, Karishma Malkan, Noah Fiedel, Noam Shazeer, Zhenzhong Lan, Yanqi Zhou, Wei Li, Nan Ding, Jake Marcus, Adam Roberts, Colin Raffel

  • Google

投稿日付(yyyy/MM/dd)

2021/2/23

概要

新規性・差分

手法

結果

コメント

@icoxfog417
Copy link
Member Author

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant