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

更充分实验,与Adam的实验效果进行比较 #21

Open
yangjianxin1 opened this issue Jun 26, 2023 · 6 comments
Open

更充分实验,与Adam的实验效果进行比较 #21

yangjianxin1 opened this issue Jun 26, 2023 · 6 comments

Comments

@yangjianxin1
Copy link

论文中比较了zero shot、lora和lomo的实验效果,但是缺少adam或者adamw的实验效果。
请问lomo和adam之间的差距有多大,是否有进行实验,期待你的回复,感谢~

@QipengGuo
Copy link
Collaborator

感谢您的问题,很有道理。但鉴于我们计算资源不是很充裕,adam相关的对比实验做得比较少(需要很大的显存来完成对比实验)。我们有在7B上做过一些小规模的测试,但不是很充分。基本是comparable的,具体在不同任务和不同数据上都会有差别。

@yangjianxin1
Copy link
Author

感谢你的回复,请问lomo和adam的对比实验,后续是否会进行补充。
如果受限于硬件资源,可以考虑在1b或者3b的模型上做快速实验,实验结论应该也是有说服力的。

@yangjianxin1
Copy link
Author

很棒的工作,因为想尝试一下lomo,但由于目前你们的论文只与lora进行了对比,所以想要了解solid的实验结果。

@QipengGuo
Copy link
Collaborator

和Adam的比较在我们的后续计划上,不过具体时间不好承诺。

@yangjianxin1
Copy link
Author

感谢回复,期待你们的更新~

@KaiLv69
Copy link
Collaborator

KaiLv69 commented Oct 20, 2023

感谢回复,期待你们的更新~

你好,可以参考这篇新论文 https://arxiv.org/pdf/2310.10195.pdf :)

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

No branches or pull requests

3 participants