You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The root cause is that the n_processes settings are not propagated to score.py:train_RF(). Thus, the default value n_jobs: int = -1 is always used there, which makes GridSearchCV (according to doc) always use the maximum number of cores available. If this is bigger than 60, the crash described in #434 occurs.
Suggested solution:
pass n_processes settings to score.py:train_RF().
(optional) to make train_RF() itself robust against this issue, introduce another cpu core check (similar to the solution in 7e989bc)
Describe the bug
The error described in #434 still occurs on machines with > 60 cores, regardless of the
n_processes
settings.The text was updated successfully, but these errors were encountered: