The Cerebros package is an ultra-precise Neural Architecture Search (NAS) / AutoML that is intended to much more closely mimic biological neurons than conventional neural network architecture strategies.
Made some code changes to support v2.16 forward compatibility (cast target variables as float32)
We are unable to upgrade to 2.16 yet until we make some elaborate changes to how we are loading base models off of of tf hub for the cifar10 and ham-spam bert encoding CICD test in the automerge workflow. With that said, both of these will probably be replaced with better examples anyway.
We are unable to upgrade to 2.16 yet until we make some elaborate changes to how we are loading base models off of of tf hub for the cifar10 and ham-spam bert encoding CICD test in the automerge workflow. With that said, both of these will probably be replaced with better examples anyway.
@sashakolpakov, do you want to merge this one?