Model's pretrained weight

[YHYeooooong]

Hi, thanks for sharing great work!

I want to know which datasets are used to learn the given model weights. In the case of ResNet101, I can find the original code of the model in the Timm document, and it was pre-trained using ImageNet 1k. But I can not find the dataset information for pre-training of 'mobilenetv2_100', 'efficient_b0', 'resmlp_24_224', 'gmlp_s16_224', 'mixer_b16_224', 'beit_base_patch16_224', and 'vit_base_patch16_224'/ I think that those models were pre-trained based on the ImaeNet-1k, but I want to know the dataset name correctly.

[rwightman]

A bit out of date, but this is the best there is for that right now
https://github.com/rwightman/pytorch-image-models/blob/master/results/model_metadata-in1k.csv

[YHYeooooong]

Thanks for the fast answer! So I found models in the list, but I can not find some of them.
Here is the list of What I want to know pre-trained dataset exactly.

mixer_b16_224 = in1k
beit_base_patch16_224 = in21k-selfsl
vit_base_patch16_224 = in21k
(2) vit_base_patch1_224_sam = in1k
resnet101 = (None)
mobilenetv2_100 = in1k
densenet121 = in1k
efficientnet_b0 = in1k
gmlp_s16_224 = in1k
resmlp_24_224 = in1k

mixer_b16_224_in21k = (None)
beit_base_patch16_224_in22k = (None)
vit_base_patch16_224_in21k = (None)

But there is some missing part of the model information. Can you give me the information on the dataset for pre-training the above models list?