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

Fine-tuning Happens in Tiny Subspaces: Exploring Intrinsic Task-specific Subspaces of Pre-trained Language Models #2120

Open
icoxfog417 opened this issue Nov 9, 2023 · 0 comments

Comments

@icoxfog417
Copy link
Member

icoxfog417 commented Nov 9, 2023

一言でいうと

大規模言語モデルの転移学習がモデル内のサブ空間で行われている可能性を示唆した研究。モデルの重みをエポックごと保存し、Flatten したうえで時系列にスタックし SVD をかけることで Fine Tuning 中のパラメーター変動をよく説明する軸を発見。この軸上で異常値を示すパラメーターを無効化すると性能が著しく下がることを確認。

論文リンク

https://arxiv.org/abs/2305.17446

著者/所属機関

Zhong Zhang, Bang Liu, Junming Shao

  • University of Electronic Science and Technology of China
  • Shenzhen Institute for Advanced Study
  • Mila & Université de Montréal

投稿日付(yyyy/MM/dd)

2023/5/27

概要

新規性・差分

手法

結果

コメント

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