search

mhamilton723 / STEGO

Unsupervised Semantic Segmentation by Distilling Feature Correspondences
MIT License
711 stars 142 forks source link

Precompute KNNS Error #64

Closed Holmes2002 closed 1 year ago

Holmes2002 commented 1 year ago

Hi I train with my dataset. My dataset has 9 classes but I adjust in topk =7 why it still error ? RuntimeError: selected index k out of range

mhamilton723 commented 1 year ago

Im not quite sure i have enough info to give you a good fix here. Could you provide more info?

Holmes2002 commented 1 year ago

I solved it. Tks for your cmt !

krishnaadithya commented 1 year ago

@Holmes2002 how did you fix this?

I am trying to train with my custom dataset and the classes =5

I get his error: all_nns.append(torch.topk(pairwise_sims, 30)[1]) RuntimeError: selected index k out of range

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

yukun80 commented 4 months ago

@mhamilton723 @krishnaadithya I have the same problem, is the parameter setting of topk wrong? Looking forward to your reply