Skip to content

Release the Model Weights of Unitok-base #27

Description

@an287-git

Dear authors,
Thank you for your excellent work on UniTok.
I was wondering if you could kindly release the pre-trained model weights for unitok-base.
I attempted to train it from scratch on my datasets, but the performance was suboptimal.
Having access to the original weights would be very helpful for further research.

Activity

  1. machuofan commented on Oct 24, 2025

    @machuofan
    Collaborator

    Hi there,

    Thank you for your interest in our work. We are sorry that the UniTok-Base model was used solely for ablation studies in our work and was not pretrained on billion-scale data. For your case, you might consider distilling the UniTok-Large model into a base model using your own dataset.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

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