resnst configs bug:
when I use resnest, I meet the problem as follow:
val_dataloader, val_cfg, and val_evaluator should be either all None or not None, but got val_dataloader
I fix it by adding the following statement in resnest config(like resnest50_32xb64_in1k.py):
val_cfg = dict()
test_cfg = dict()
Branch
main branch (mmpretrain version)
Describe the bug
resnst configs bug: when I use resnest, I meet the problem as follow: val_dataloader, val_cfg, and val_evaluator should be either all None or not None, but got val_dataloader
I fix it by adding the following statement in resnest config(like resnest50_32xb64_in1k.py): val_cfg = dict() test_cfg = dict()
Environment
python 3.9.16 h6244533_2 pytorch 1.13.1 py3.9_cuda11.6_cudnn8_0 pytorch pytorch-cuda 11.6 h867d48c_1 pytorch torchaudio 0.13.1 pypi_0 pypi torchvision 0.14.1 pypi_0 pypi mmcv 2.0.0 pypi_0 pypi mmengine 0.7.4 pypi_0 pypi mmpretrain 1.0.0rc8 dev_0
Other information
No response