Skip to content

Integrate DeepSpeed  #22

@simi2525

Description

@simi2525

Integrate Microsoft's deepspeed into the template https://www.deepspeed.ai/
This will allow for the usage of arbitrarily bigger models which don't have to fit in VRAM.

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