Mingzhou Liu
Mingzhou Liu
@renero Hi, could you please recommend some Python/R implementations for FCI and RFCI? Thanks a lot.
@shahdhardn Sorry, I didn't find a solution for this problem. I used CPU version cdt docker instead. I am waiting the authors of cdt to offer a solution.
> I meet same issue, but I have solved it when I changed the learn rate to 0.0001. I hope it can help you. It works for me , too...
@eypros Hi~ About the penultimate number , I notice they have different intervals when TotalFrameNumber changes. Do you know how the penultimate number are set(calculated)? Thanks!
I set the 'is_train' to False, but I find the size of 'test_image.png' is 336*336 and it seems the test image resolution is very low . Can anyone help me?
@FancccyRay @lfz Have you solve this problem? I notice a relative answer in #21 .When shoul we use clip_grad_norm?
@hflyzju I encounted the same problem as yours. I changed b2 from 12 to 16(2*N), then got a stage-2 AUC=0.82
```\dt *.*``` works for me.
Setting the learning rate to 1e-3 and cancelling the lr_decrease helps in my case.