Skip to content

Clarification regarding dropout in the multihead attention block #21

@anupsingh15

Description

@anupsingh15

Hi @hkproj

Why do you add dropout to the attention scores (line 110 in model.py)? Shouldn't you discard the dropout in the multihead attention block because you already add a dropout (line 81) in the residual connection block?

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