search

MrChenFeng / SSR_BMVC2022

SSR: An Efficient and Robust Framework for Learning with Unknown Label Noise (BMVC2022)
https://bmvc2022.mpi-inf.mpg.de/372/
MIT License
29 stars 7 forks source link

Issues trying to re-run the code #6

Closed AlejandraM97 closed 8 months ago

AlejandraM97 commented 8 months ago

Hi!

Hope you are well, I've seen your paper, and it is pretty interesting... I'm trying to re-run it... nevertheless, this problem shows

Epoch [0/300] selection: theta_s:1.0 TP: 4561 FP:150 TN:9754 FN:35535 Epoch [0/300] relabelling: correct: 0 original: 0 total: 0 0%| | 0/390 [00:00<?, ?it/s]Traceback (most recent call last): File "/data/arumota/SSR_BMVC2022/main_cifar.py", line 62, in train [inputs_x1, inputs_x2], labelsx, , index = labeled_train_iter.next() AttributeError: '_MultiProcessingDataLoaderIter' object has no attribute 'next'

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/data/arumota/SSR_BMVC2022/main_cifar.py", line 316, in main() File "/data/arumota/SSR_BMVC2022/main_cifar.py", line 287, in main train(labeled_loader, modified_label, all_loader, encoder, classifier, proj_head, pred_head, optimizer, i, args) File "/data/arumota/SSR_BMVC2022/main_cifar.py", line 65, in train [inputs_x1, inputs_x2], labelsx, , index = labeled_train_iter.next() AttributeError: '_MultiProcessingDataLoaderIter' object has no attribute 'next'

I know it is because of the dataset... how do u process this data?

MrChenFeng commented 8 months ago

.next() is not supported by new torch, I have updated the code.