RuntimeError: Expected object of type Variable but found type CUDAType for argument #0 'result'

[xjiao004]

Anyone can help me with this problem here? This is the logs. Really appreciate it.

RuntimeError: Expected object of type Variable but found type CUDAType for argument #0 'result' (checked_cast_variable at /opt/conda/conda-bld/pytorch_1556653099582/work/torch/csrc/autograd/VariableTypeManual.cpp:173)
frame #0: c10::Error::Error(c10::SourceLocation, std::string const&) + 0x45 (0x7f270e481dc5 in /home/supernode/anaconda3/envs/CenterNet/lib/python3.6/site-packages/torch/lib/libc10.so)
frame #1: torch::autograd::VariableType::checked_cast_variable(at::Tensor&, char const*, int) + 0x169 .............................................

[li10141110]

same problem, pytorch 1.1.0 cuda 9 and nvidia gtx 2080ti

[xjiao004]

Same environment, it's the torch version's problem. Try pytorch=0.4.1 with cuda9.0

[li10141110]

Torch version bug but i need torch1.1.0 environment, any help

[hejinsome]

@xjiao004 我是titan rtx，采用pytorch1.1和cuda10出问题，采用pytorch1.0与cuda9也出问题

[Duankaiwen]

@hejinsome princeton-vl/CornerNet@3809432

[mama110]

@li10141110, have you solved the problem?

[YIGAOYI]

Using the _cpools layer in CornerNet-Lite works for me. Just replace it and setup again.
https://github.com/princeton-vl/CornerNet-Lite/tree/master/core/models/py_utils/_cpools
Thanks! pytorch 1.1.0 cuda 10