search

NVlabs / FB-BEV

Official PyTorch implementation of FB-BEV & FB-OCC - Forward-backward view transformation for vision-centric autonomous driving perception
Other
608 stars 40 forks source link

Why does training with 8 cards report an error? #40

Open beizhengren opened 1 month ago

beizhengren commented 1 month ago

The detail of the error is:

mmdet3d/datasets/samplers/infinite_group_each_sample_in_batch_sampler.py", line 174, 
assert self.groups_num >= self.global_batch_size

The error message is located on line 174, which belongs to the initialization function of the InfiniteGroupEachSampleInBatchSamplerEval.

and the print info of relative variable is:

batch_size :  1,
world_size: 1,
global_batch_size: 1,
groups_num: 2

How can I slove the problem, Thanks!