Skip to content

How to Run Code on Multiple Machines #47

@LuoMayao

Description

@LuoMayao

Great work, but I think I've encountered some issues. I tried using the torchrun command to train DIT across multiple machines, but it seems unable to correctly detect the rank. Could you please provide the appropriate multi-machine launch command that the code is compatible with?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions