Skip to content

[Question]: chatglm-6b，在微调的时候，loss和 rouge得分一般达到多少就算模型比较好了？ #5934

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.

Sign up for GitHub

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

Jump to bottom
Closed
liuzhipengchd opened this issue May 17, 2023 · 1 comment
Closed

[Question]: chatglm-6b，在微调的时候，loss和 rouge得分一般达到多少就算模型比较好了？ #5934

liuzhipengchd opened this issue May 17, 2023 · 1 comment
Assignees
@wawltor
Labels
question Further information is requested triage

Comments

@liuzhipengchd
Copy link

请提出你的问题

我使用的是自己的数据，一问一答这种类型数据。。prompt的文本比较短而且重复性较高，answer比较长。在实际微调的时候，loss很难降低 rouge都是0.1多
@liuzhipengchd liuzhipengchd added the question Further information is requested label May 17, 2023
@w5688414
Copy link
Contributor

w5688414 commented May 8, 2024

可以使用lora进行训练，具体loss和rouge与数据集有关，可以构建典型的评估集进行评估。

https://github.com/PaddlePaddle/PaddleNLP/tree/develop/llm

@paddle-bot paddle-bot bot closed this as completed May 13, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@wawltor wawltor

Labels
question Further information is requested triage
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@w5688414 @wawltor @liuzhipengchd