单个epoch多个batch之间loss抖动过大 #6

Open

opened

on Aug 13, 2021

使用作者代码训练数据，在模型收敛之后，loss抖动仍然很大，和其它代码相比则无此类问题（在英语训练语料上）

to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Assignees

No one assigned

Labels

No labels

No labels

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests