Unanswered
Hi All, I Am Trying To Execute Somewhat Custom Hpo Scheme With Clearml.
I Would Want That A Single Running Python Script Will Be Able To Sample The Optimizer, Init A Task And Report The Result Multiple Times. I Didn'T Find Anything Similar In The Docs Or
But it does make me think, if instead of changing the optimizer I launch a few workers that "pull" enqueued tasks, and then report values for them in such a way that the optimizer is triggered to collect the results? would it be possible?
But this is Exactly how the optimizer works.
Regardless of the optimizer (OptimizerOptuna or OptimizerBOHB) both set the next step based on the scalars reported by the tasks executed by agents (on remote machines), then decide on the next set of parameters in a Bayesian manner. What am I missing here ?
152 Views
0
Answers
3 years ago
one year ago