Add share_optimizer option on CmaEsSampler#1776
Add share_optimizer option on CmaEsSampler#1776c-bata wants to merge 5 commits intooptuna:masterfrom
Conversation
|
@HideakiImamura @c-bata Thanks for this! Is there an easy way for me to install this version locally to verify that it fixes the issue? |
|
@maartenpants You can install this version from the @c-bata's specific branch by |
|
@c-bata We just encountered the same error again while running with this branch. |
|
It only seems to occur after many epochs. CC @HideakiImamura |
|
Hi @maartenpants! Thank you for checking this branch. study = optuna.create_study(
storage=...,
sampler=optuna.samplers.CmaEsSampler(share_optimizer=False),
) |
|
@c-bata Ah! Thanks |
|
Will this have any effect on the sampling process? We're running with 4 "workers" to speed up the optimization. |
Yes. |
|
I opened #1833 for an alternative solution. It has no effects on the sampling process. |
Motivation
fix #1775
Description of the changes
share_optimizeroption which enables to store CMA object in system_attrs (default: False).randomize_x0option to sample initial starting points randomly.Discussion points
Which ones should be set as a default argument,
share_optimizer=Falseorshare_optimizer=True?share_optimizer=Falsefrom the next release: Existing users may be confused by this breaking change. From the next release, CmaEsSampler cannot resume optimizations.share_optimizer=Trueuntil v3.0.0 release: It keeps current behavior. New users may be confused because this option does not work with PostgreSQL and MySQL.