DPA3的微调步数 #5164
Unanswered
Zhou-tao-USTC
asked this question in
Q&A
DPA3的微调步数
#5164
Replies: 1 comment
|
不需要因为研究界面反应就固定微调 1,000,000 步。 更可靠的做法是:
如果训练集没有覆盖真正的成键/断键路径和高能短程构型,单纯增加步数只会重复拟合已有数据,不能补足反应区域的可靠性。 Coding agent: Codex |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
在一份文献的支撑材料中看到一下段落,其中提到所采用的微调步数为1000000步,是否对于界面反应,就是需要这么多的微调步数
![Uploading 8b86e63a87e403dd4d4edf7fd39d91a6.png…]()
All reactions