search

HeliosZhao / M3L

Pytorch implementation for M^3L. CVPR 2021
65 stars 21 forks source link

question after my run #4

Closed AnabellaHan closed 2 years ago

AnabellaHan commented 3 years ago

I ran your project with three GPUs,and i didnot change the code except replacing msmt17v1 with msmt17v2,but it has this question:

File "D:\papercode\M3Lmaster\reid\trainers.py", line 64, in train f_out, tri_features = self.model(inputs, MTE='', save_index=save_index) ValueError: too many values to unpack (expected 2)

Is there something wrong with my dataset? Have you encountered it before? Thank you for reading!

HeliosZhao commented 2 years ago

Hi, From your traceback, it seems that it is not a problem about dataset. Do you add --BNNect when running?

AnabellaHan commented 2 years ago

My problem has been solved,Thank you!

Cherishhuier commented 1 week ago

Hello, I get the following error when I run your code again, is it because there is a problem with my dataset? Have you encountered such a problem? Thank you very much for visiting. /M3L-master/reid/utils/data/preprocessor.py", line 27, in _get_single_item fname, pid, camid, dataset_id = self.dataset[index] ValueError: not enough values to unpack (expected 4, got 3)

Cherishhuier commented 1 week ago

Although the time is a bit long, as a new scientific research, I hope to get your answer, thank you very much.