VAD Train Problem

[Yjerry001]

In Step 2 "Prepare Bench2Drive data info" of the Prepare Bench2Drive Dataset, I reduced the maps due to memory limitations.
However, when running VAD Train
./adzoo/vad/dist_train.sh ./adzoo/vad/configs/VAD/VAD_base_e2e_b2d.py 1
I encountered the following error, which indicates that the dataset length is 0. Do you have any suggestions? Where should I check if the dataset was not successfully loaded? Thank you!

我在Prepare Bench2Drive Dataset的步骤2Prepare Bench2Drive data info中,由于内存原因,将地图减少了.
然后在运行VAD Train
./adzoo/vad/dist_train.sh ./adzoo/vad/configs/VAD/VAD_base_e2e_b2d.py 1
的过程中,遇到如下报错,显示dataset长度为0,请问有没有什么建议,dataset没有成功传入应该检查哪里,谢谢!

[rank0]: Traceback (most recent call last):
[rank0]:   File "./adzoo/vad/train.py", line 237, in <module>
[rank0]:     main()
[rank0]:   File "./adzoo/vad/train.py", line 226, in main
[rank0]:     custom_train_model(
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/adzoo/bevformer/mmdet3d_plugin/bevformer/apis/train.py", line 25, in custom_train_model
[rank0]:     custom_train_detector(
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/adzoo/bevformer/mmdet3d_plugin/bevformer/apis/mmdet_train.py", line 206, in custom_train_detector
[rank0]:     runner.run(data_loaders, cfg.workflow)
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/runner/epoch_based_runner.py", line 129, in run
[rank0]:     epoch_runner(data_loaders[i], **kwargs)
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/runner/epoch_based_runner.py", line 55, in train
[rank0]:     self.call_hook('after_train_epoch')
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/runner/base_runner.py", line 297, in call_hook
[rank0]:     getattr(hook, fn_name)(self)
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/runner/hooks/evaluation.py", line 266, in after_train_epoch
[rank0]:     self._do_evaluate(runner)
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/core/evaluation/eval_hooks.py", line 121, in _do_evaluate
[rank0]:     results = self.test_fn(
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/adzoo/bevformer/apis/test.py", line 69, in custom_multi_gpu_test
[rank0]:     for i, data in enumerate(data_loader):
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 440, in __iter__
[rank0]:     return self._get_iterator()
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 388, in _get_iterator
[rank0]:     return _MultiProcessingDataLoaderIter(self)
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1083, in __init__
[rank0]:     self._reset(loader, first_iter=True)
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1116, in _reset
[rank0]:     self._try_put_index()
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1350, in _try_put_index
[rank0]:     index = self._next_index()
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 620, in _next_index
[rank0]:     return next(self._sampler_iter)  # may raise StopIteration
[rank0]:   File "/home/yjr/.local/lib/python3.8/site-packages/torch/utils/data/sampler.py", line 288, in __iter__
[rank0]:     for idx in self.sampler:
[rank0]:   File "/home/yjr/Bench2DriveZoo/Bench2DriveZoo-uniad-vad/mmcv/datasets/samplers/distributed_sampler.py", line 32, in __iter__
[rank0]:     math.ceil(self.total_size / len(indices)))[:self.total_size]
[rank0]: ZeroDivisionError: division by zero

[jiaxiaosong1002]

@Yjerry001 Not sure about the implications of reducing maps.