Skip to content

Dfine nano version learning rate config error #132

@qwuzer

Description

@qwuzer

## Our LR-Scheduler
flat_epoch: 7800 # 4 + epoch // 2, e.g., 40 = 4 + 72 / 2
no_aug_epoch: 12
lr_gamma: 1.0

In this nano version config, the flat epoch seems to be set wrong to 7800, which should be 78 instead. Also, why is lr_gamma set as 1.0, that is strange since wouldn't that cause the learning rate to be the same? Should I modify that to 0.5 so it match the other versions' base lr in

lr_gamma: 0.5
with 0.5?

Have been trying to train a DEIM model that works better on long tail objects for my machine learning course, if anyone could help answer my question here would be a great help, thanks!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    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