Checklist

• I have searched related issues but cannot get the expected help.
• I have read related documents and don't know what to do.

Describe the question you meet

hi developer, when I try to change backbone resnet of teachers and students to swin

I got the following error
Traceback (most recent call last): File "./tools/mmcls/train_mmcls.py", line 203, in <module> main() File "./tools/mmcls/train_mmcls.py", line 174, in main algorithm.init_weights() File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/base_module.py", line 116, in init_weights m.init_weights() File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/base_module.py", line 116, in init_weights m.init_weights() File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/base_module.py", line 105, in init_weights initialize(self, self.init_cfg) File "/opt/conda/lib/python3.7/site-packages/mmcv/cnn/utils/weight_init.py", line 613, in initialize _initialize(module, cp_cfg) File "/opt/conda/lib/python3.7/site-packages/mmcv/cnn/utils/weight_init.py", line 517, in _initialize func(module) File "/opt/conda/lib/python3.7/site-packages/mmcv/cnn/utils/weight_init.py", line 494, in __call__ logger=logger) File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/checkpoint.py", line 602, in load_checkpoint load_state_dict(model, state_dict, strict, logger) File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/checkpoint.py", line 80, in load_state_dict load(module) File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/checkpoint.py", line 78, in load load(child, prefix + name + '.') File "/opt/conda/lib/python3.7/site-packages/mmcv/runner/checkpoint.py", line 75, in load err_msg) File "/opt/conda/lib/python3.7/site-packages/mmcls-0.23.2-py3.7.egg/mmcls/models/backbones/swin_transformer.py", line 457, in _load_from_state_dict *args, **kwargs) File "/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1127, in _load_from_state_dict hook(state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs) File "/opt/conda/lib/python3.7/site-packages/mmcls-0.23.2-py3.7.egg/mmcls/models/backbones/swin_transformer.py", line 525, in _prepare_relative_position_bias_table relative_position_bias_table_pretrained, nH1) File "/opt/conda/lib/python3.7/site-packages/mmcls-0.23.2-py3.7.egg/mmcls/models/utils/embed.py", line 104, in resize_relative_position_bias_table z = table[:, i].view(src_shape, src_shape).float().numpy() TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

the command I use is
python ./tools/mmcls/train_mmcls.py \ /workspace/mmrazor/configs/distill/wsld/swin_tiny_distill_wsld_01.py \ --work-dir ./work_dir \ --cfg-options algorithm.distiller.teacher.init_cfg.type=Pretrained