Extremely slow if I tune rsm (mtry)
Open
performance
R
- Dominant language
- C++
- Stars
- 9.1k
- Forks
- 1.3k
- PR merge metrics
- No merged PRs in 30d
Description
Problem:
I'm tuning the hyperparameters (using R) and even though everything is working just fine, I almost never tune rsm since it's extremely slow. What is the reason for this?
catboost version:
0.26.1
Operating System:
Windows 10
CPU:
Intel(R) Core(TM) i5-8350U CPU @ 1.70GHz 1.90 GHz
This has also happend on quite larger virtual machines with other catboost versions (on Windows).
Contributor guide
Assessment
This issue has not been assessed yet.