Skip to content

solver stops early on non-convex optimization #100

@jeisner

Description

@jeisner

Trying lesson 9 with L2 regularization, C=-3 or C=-30. The solver seems to stop after a few iterations when the gradient is not 0. We should still climb to a local maximum (possibly with some infinite weights), no?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions