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.
Ultimately, make the selection of Adam, AdamW, Lion a tunable param (they both have learningrate as a param. Will need an integer selection as a key referencing which one, as Katib can't directly select a class.)
Kind of issue: feature-request-or-enhancement
Add to the API:
In the stable API:
The tf.keras.optimizers.AdamW
To the experimental API add the Lion optimizer:
https://github.com/GLambard/Lion-tensorflow
Ultimately, make the selection of Adam, AdamW, Lion a tunable param (they both have learningrate as a param. Will need an integer selection as a key referencing which one, as Katib can't directly select a class.)