search

zkyseu / O2SFormer

End-to-End Lane detection with One to Several Transformer
Apache License 2.0
35 stars 9 forks source link

ValueError: cannot convert float NaN to integer #25

Open hard-workingXu opened 3 months ago

hard-workingXu commented 3 months ago

Hello, I ran into this problem in training, it looks like I can't convert the floating point NaN (null) to an integer, is this problem the data itself? Or is it something else? The following are the specific error reports： fatal: not a git repository (or any parent up to mount point /) Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set). /home/lx/anaconda3/envs/0s2/lib/python3.7/site-packages/mmdet/datasets/samplers/group_sampler.py:21: RuntimeWarning: invalid value encountered in long_scalars size / self.samples_per_gpu)) self.samples_per_gpu Traceback (most recent call last): File "train_net.py", line 248, in main() File "train_net.py", line 244, in main meta=meta) File "/home/lx/桌面/O2SFormer-main/dnlane/apis/train.py", line 154, in train_detector data_loaders = [build_dataloader(ds, train_loader_cfg) for ds in dataset] File "/home/lx/桌面/O2SFormer-main/dnlane/apis/train.py", line 154, in data_loaders = [build_dataloader(ds, train_loader_cfg) for ds in dataset] File "/home/lx/anaconda3/envs/0s2/lib/python3.7/site-packages/mmdet/datasets/builder.py", line 181, in build_dataloader samples_per_gpu) if shuffle else None File "/home/lx/anaconda3/envs/0s2/lib/python3.7/site-packages/mmdet/datasets/samplers/group_sampler.py", line 21, in init size / self.samples_per_gpu)) self.samples_per_gpu ValueError: cannot convert float NaN to integer

zkyredstart commented 2 months ago

Hi, what is the version of your mmdet and mmcv?