optuna / kurobako-py

A Python library to help implement kurobako's solvers and problems
MIT License
9 stars 7 forks source link

Add `warm_starting_trials` option to `OptunaSolverFactory` #9

Closed sile closed 3 years ago

sile commented 3 years ago

Until the number of trials exceeds the option value, they are evaluated with zero cost (i.e., step==0).

Note that setting a value greater than 0, you need to use a problem that supports warm-staring (i.e., WarmStartingProblem introduced by https://github.com/sile/kurobako/pull/36).