search

ExplainableML / czsl

PyTorch CZSL framework containing GQA, the open-world setting, and the CGE and CompCos methods.
GNU General Public License v3.0
111 stars 27 forks source link

Qualitative results. #24

Closed jiayaohua closed 1 year ago

jiayaohua commented 2 years ago

How do you check the qualitative results in your paper?

mancinimassimiliano commented 1 year ago

Hello @jiayaohua and sorry for the very late reply!

If you want to check qualitatives, you should like predictions of the model to the actual images in the datasets. You can do it by enabling return_images=True (see this) while creating the dataset (e.g. here). The image path will then be added at the end of your data list, and you can use data[-1] to check the path (see here even if for now it is implemented only for val/test splits).

Hope the answer is still useful, but please feel free to ask questions in case of doubts.