Christian Bartz
Christian Bartz
Are you sure you are using CUDA 8.0 on your machine?
Well, then I don't know... I did not ever use cupy in Version 6, yet... So this might be an issue. Did you try to use the docker container?
> Does it mean that when creating the gt file with --max-words 3, you should also use the --min-words 3 and create a gt file with only images containing exactly...
> It might help to swap point 3 and 4 in the corresponding readme part to avoid confusion. thanks for the hint! I changed it in the `README`. Yes, I...
Thanks for your feedback! > I've found another likely issue in the code. Yes, you are right. This definitely is an issue in the code (it also explains why it...
> I'll see if I can figure out a better way to stop it. Yeah, the code raises `StopIteration` to signal the end. You could just add a `try: except:`...
Wow, good catch. You are totally right. This way of splitting is definitely not a good idea. I think the code is still like this, because I never really really...
> I can check the length of the splits the same way as you do in `train_fsns.py`, right? Exactly :smile: Thanks for having a look at the code and playing...
Hmm, interesting error. I can think of two different causes: 1. BatchNorm sometimes has problems when run with CuDNN, but I haven't had this problem in quite a while. It...
Honestly, I can't remember. That is too long ago. I think I always used the fast validator during training because the other just needed to much time. It could be...