RuntimeError: cuDNN error: CUDNN_STATUS_MAPPING_ERROR

[dollychu]

Traceback (most recent call last): File "tools/train.py", line 224, in confmap_preds = rodnet(data.float().cuda()) File "/home/dolly/anaconda3/envs/rodnet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 532, in call result = self.forward(*input, kwargs) File "/home/dolly/Desktop/RODNet/rodnet/models/rodnet_cdc.py", line 12, in forward x = self.cdc(x) File "/home/dolly/anaconda3/envs/rodnet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 532, in call result = self.forward(*input, *kwargs) File "/home/dolly/Desktop/RODNet/rodnet/models/backbones/cdc.py", line 14, in forward x = self.encoder(x) File "/home/dolly/anaconda3/envs/rodnet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 532, in call result = self.forward(input, kwargs) File "/home/dolly/Desktop/RODNet/rodnet/models/backbones/cdc.py", line 47, in forward x = self.conv1a(x) # (B, 2, W, 128, 128) -> (B, 64, W, 128, 128) File "/home/dolly/anaconda3/envs/rodnet/lib/python3.8/site-packages/torch/nn/modules/module.py", line 532, in call result = self.forward(*input, **kwargs) File "/home/dolly/anaconda3/envs/rodnet/lib/python3.8/site-packages/torch/nn/modules/conv.py", line 479, in forward return F.conv3d(input, self.weight, self.bias, self.stride, RuntimeError: cuDNN error: CUDNN_STATUS_MAPPING_ERROR

I have followed the instructions but it keeps throwing this error message to me, does anyone facing the same problem?

[fumgmin]

I also encountered this problem, please ask which version of cuda you are using, I use cuda11.3, A6000 graphics card, currently suspect that the version is too high to match

[promaxj]

@dollychu What GPU do you use?