8000 Can I use TGI in a Supercomputer? · Issue #3208 · huggingface/text-generation-inference · GitHub
[go: up one dir, main page]

Skip to content
Can I use TGI in a Supercomputer? #3208
Open
@sleepingcat4

Description

@sleepingcat4

I want to generate somewhere around 1 trillion tokens and I was thinking of using TGI on a European Supercomputer. is there a way to achieve this without relying on docker and downloading the model natively and then load it on the compute node and serve it? @Wauplin

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

      0