Skip to content

Adding "SmartGradient" Approach for the Hyperparameter Gradient Computation #103

@vincent-maillou

Description

@vincent-maillou

Currently, the gradient of the hyperparameters is computed along the "default axis", one can use the last gradient descent direction to create an orthonormal basis and compute the next gradient along these directions.
This is referred to as the "SmartGradient" approach and is described here: https://arxiv.org/abs/2106.07313

Metadata

Metadata

Assignees

No one assigned

    Labels

    New FeatureThe issue relate to a new issue wanted in the framework.Performance improvementThe issue relate to runtime and/or memory performance improvement.

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions