Skip to content

Has the model weight issue been resolved? #14

@ludandandan

Description

@ludandandan

Thank you for your work. I would like to know if the model weights have been updated. The results I calculated using the current model weights are still different from those in Table 1 of the paper (even lower than the CLIP ViT-L results reported in the paper). The results obtained using CLIP ViT-L are the same as those reported in Table 1 of the paper.
The results of CLIP ViT-L are as follows

Image

The results of CSD are as follows

Image

Table 1

Image

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions