Where can I find the file 'rn18_bn0.01_[4K]_x_l2_x_tv.crop'?

VILA-Lab / SRe2L

(NeurIPS 2023 spotlight) Large-scale Dataset Distillation/Condensation, 50 IPC (Images Per Class) achieves the highest 60.8% on original ImageNet-1K val set.

119 stars 16 forks source link

Where can I find the file 'rn18_bn0.01_[4K]_x_l2_x_tv.crop'? #3

Closed muyangli17 closed 1 year ago

muyangli17 commented 1 year ago

I am trying running train_FKD.py with train.sh, but I did not find the file rn18_bn0.01_[4K]_x_l2_x_tv.crop for argument --train-dir in the dataset link(https://mbzuaiac-my.sharepoint.com/personal/zeyuan_yin_mbzuai_ac_ae/_layouts/15/onedrive.aspx?id=%2Fpersonal%2Fzeyuan%5Fyin%5Fmbzuai%5Fac%5Fae%2FDocuments%2Fproject%5Fshare%2FSRe2L&ga=1). Can you show me where I can find it? Thank you zeyuanyin.

muyangli17 commented 1 year ago

Sorry, I mistook the procedure. Such file will be created from the procedure _datasynthesis.py, right?

zeyuanyin commented 1 year ago

For detailed dataset information, please refer to Download section of our project page at https://zeyuanyin.github.io/projects/SRe2L/#Download.

You can use our provided synthetic datasets to reproduce our results or you can generate them step by step.