Problem about "Was asked to gather along dimension 0"

[Dong ZHANG]

I met the following warning when I run bash ./launch/run_voc_resnet38.sh on 8 GPUs. I wonder if this warning would affect model performance?

"anaconda3/lib/python3.6/site-packages/torch/nn/parallel/_functions.py:61: UserWarning: Was asked to gather along dimension 0, but all input tensors were scalars; will instead unsqueeze and return a vector.
warnings.warn('Was asked to gather along dimension 0, but all '"

Thanks.

[Nikita Araslanov]

Hi @ZHANGDONG-NJUST,
I don't think it should be the issue. If I understand correctly, it's probably because the losses are returned from .forward() as scalars, and not as tensors of shape (1, ).
Best,
Nikita

[Dong ZHANG]

Thanks for your reply. Do you think this may affect the model performance ?