nip.parameters.types#
Enum types for experiment parameters.
Used when a parameter can only take on a specific set of values.
Module Attributes
Type for the scenario to run. |
|
Type for SPG variants. |
|
Type for the variants of the inverse Hessian-vector product computation. |
|
Type for the possible guesses of the verifier in binary cases. |
|
Type for the RL trainer to use. |
|
Enum specifying on which iterations to test the model during training. |
|
Type for the PPO loss function to use. |
|
Type for ways of turning a multi-class classification task into a binary one. |
|
Type for the activation function to use. |
|
Type for the interaction protocol to use in the environment. |
|
Type for the type of building block to use in the image classification network. |
|
Type for the method to use to do regression analysis on the messages. |
|
Enum for how to base the current experiment on a previous W&B run. |