Skip to content

optimizer.iterations didn't change all the time. #2

Description

@hehongjie

I modified the train_eager.py and ran the program successfully with tensorflow 2.1, while the checkpoint was saved each iteration. As a result, those checkpoint files took up all the disk space and the training process cannot be finished successufully. Is there any missing code about optimizer.iteration? Or optimizer.iteration can only work in keras program?

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