Computer scienceData scienceMachine learningIntroduction to deep learningOptimizers

Adam

Beta change

Report a typo

How common is it to set $\beta_1$ to 0?

Select one option from the list

When

\beta_2

is set to a low value,

\beta_1

should also be set to a similarly low value

\beta_2

is set to a high value (e.g., 0.999),

\beta_1

is always set to 0

Not that common, but certain GAN-based architectures tried using this setting

___

Create a free account to access the full topic