Skip to content

Please release Medusa-fine-tuned 4B? #140

Description

@lenn-arts

Dear Cosmos Team,
Thank you for doing the community a service by releasing your models open-source. In the report you dedicate section 5.2.4 to describing how you optimized the inference time speed of Cosmos-Predict-4B to real-time with a variety of techniques including Medusa speculative decoding.
Would you consider releasing this fine-tuned model to the public too? I am confident many researchers like myself would find this extremely useful for downstream use cases that require real-time speed.
Thank you!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Fields

    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