search

hila-chefer / Transformer-Explainability

[CVPR 2021] Official PyTorch implementation for Transformer Interpretability Beyond Attention Visualization, a novel method to visualize classifications by Transformer based networks.
MIT License
1.75k stars 232 forks source link

How can I adapt this to my custom model? #58

Closed sarah-muller closed 1 year ago

sarah-muller commented 1 year ago

I understood that Transformer-Explainability can only be used with the following three models. How can I use it with my custom model?

default_cfgs = {
    # patch models
    'vit_small_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/vit_small_p16_224-15ec54c9.pth',
    ),
    'vit_base_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_p16_224-80ecf9dd.pth',
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5),
    ),
    'vit_large_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_p16_224-4ee7a4dc.pth',
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
}
hila-chefer commented 1 year ago

Hi @sarah-muller, thanks for your interest! Please see this issue