Skip to content

Cannot reproduce the results of the paper #3

Description

@zaccharieramzi

Hi @aravindr93 ,

Thank you so much for providing this repo and the code. I was able to run it in relatively little amount of time and get some results.

However, even when raising the number of tasks to 200,000 and using regularization of 2.0 in the evaluation, I get the following accuracy for validation: 98.204%, which is far from the reported 99.5% on the 5 way 1 shot case of Omniglot.

Do you know if there is any explanation to this? Do you have the weights of models trained to get 99.5% accuracy? I would be very interested in those.

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