A Code Issue About “pretrain/main.py”

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

1.41k stars 82 forks source link

A Code Issue About “pretrain/main.py” #79

Closed lynnz7 closed 6 months ago

lynnz7 commented 6 months ago

Hello, your work is very meaningful and effective. While reviewing your code, I noticed a SparK.forward statement on line 156 in SparK/pretrain/main.py. I'm not quite familiar with this usage. Is it a typo or does it serve another purpose? I hope you can clarify this. Thank you.

keyu-tian commented 6 months ago

Thank you @lynnz7. It's a useless statement and can be removed. I did this only for more hints from PyCharm.

lynnz7 commented 6 months ago

Thank you for your thoughtful response, and wishing you a Happy Chinese New Year!