How to solve the error when the clustering input value is NaN?

[UserHu202]

When I run the ACORM algorithm for a while, the clustering algorithm piece reports an error, specifically: ValueError: Input X contains NaN.KMeans does not accept missing values encoded as NaN natively.

[qqgs]

Hello, I have the same problem, did you solve it? How was it solved?

[UserHu202]

Hello, I have the same problem, did you solve it? How was it solved?
The gradient exploded, and running it with the parameter in the author's run.sh solves the problem

[huzican]

If the parameter of main.py may lead to this problem, please run with run.sh file, or fill the parameter of run.sh into main.py.