hjimce / o2u-net Goto Github PK

Hi,
thanks for sharing your implementation. I have two questions about it:

1. Does it also work on tabular data?
2. Is the code tailored to the datasets used in the paper or can one apply it to any data?

Thanks!

Is there a missing model file? Because in main.py there is: from model Import cotnet

we employ the resnet101 model and perform the experiments in the CIFAR100 dataset. We didn't make any changes to the source code. However, we can't reproduce your results.

The red part is the result of my running, which records the training results of the model in stage three. Moreover, the performance of "coteaching" is also not well.

I need to secure my number.

The pseudocode in the paper for Step 1 says "repeat...until stable accuracy and loss in the validation set", however, there is no evidence of early stopping in first_stage of main.py