why pretrained models' query result different between methods of list_models() and model_entrypoints

[Ga-Lee]

Hello! Thank you for the excellent job!
If I use the function , model_list = list_models(cswin) to check whether some models(for example, CSWin-transformer) pretrained weights are available, and I get model_list =[] (And my timm version is 0.5.4), does it mean the model and its weights are not collected by timm?

And by the way, is the model results of imagenet val (https://github.com/rwightman/pytorch-image-models/blob/master/results/results-imagenet.csv) is up-to-data ？ for new models added into timm, are their performance on imagenet1k included into the csv?

Thank you for your reply, you kindness will help a great lot!

[rwightman]

@Ga-Lee cswin isn't included in timm. the csv files are updated every few months so it can lag behind the current master, but usually at or ahead of the last official pypi release

[Ga-Lee]

Got it ! Appreciate for your reply!