René Sass

The documentation still has to be updated drastically. But you could use following examples already: https://github.com/automl/SMAC3/blob/development-2.0/examples/mlp_multi_fidelity.py https://github.com/automl/SMAC3/blob/development-2.0/examples/sgd_instances.py or also the other examples but you have to set deterministic in the...

https://github.com/automl/pynisher/tree/api_changes

Will be closed by #853

Placeholder: List how old variables are called

Really weird behaviour indeed. I just removed it.

I added the option to pass a runhistory in development 2.0. The initial configs and incumbent are derived from the runhistory directly. Does that fix the problem?

I've got the same error. We will investigate this further. Thanks for pointing this out!

Hey, I actually don't see your problem here. If you use a different seed on `SMAC4AC` then also different configurations are sampled, resulting (depending how big/small your configspace is) in...

Depends. If your TAE is deterministic and you pass and utilize the seed (to your random forest) then yes. Otherwise not. Setting no seed to the random forest means that...

Remove runtime optimization?