keyu-tian / SparK

[ICLR'23 Spotlight🔥] The first successful BERT/MAE-style pretraining on any convolutional network; Pytorch impl. of "Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling"
https://arxiv.org/abs/2301.03580
MIT License
1.46k stars 84 forks source link

there is no requirements.txt file. #70

Closed mandal4 closed 11 months ago

mandal4 commented 11 months ago

Hi, thanks for releasing the source code. Could you share the requrements.txt file? It doesn't exists in this repository and i have some trouble about package version.

Thanks,

mandal4 commented 11 months ago

I found it in ./downstream_imagenet