RuntimeError: cuDNN error: CUDNN_STATUS_EXECUTION_FAILED

显卡：RTX 3080

cuda: 11.1

cudnn: 8.0.5

虚拟环境： python3.5.4   pytorch1.2.0   torchvision 0.4.0  

Traceback (most recent call last):
  File "./tools/train.py", line 240, in
    main()
  File "./tools/train.py", line 142, in main
    pred_r, pred_t, pred_c, emb = estimator(img, points, choose, idx)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/DenseFusion-Pytorch-1.0/lib/network.py", line 96, in forward
    out_img = self.cnn(img)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/DenseFusion-Pytorch-1.0/lib/network.py", line 36, in forward
    x = self.model(x)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward
    return self.module(*inputs[0], **kwargs[0])
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/DenseFusion-Pytorch-1.0/lib/pspnet.py", line 65, in forward
    f, class_f = self.feats(x) 
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/DenseFusion-Pytorch-1.0/lib/extractors.py", line 119, in forward
    x = self.layer1(x)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/container.py", line 92, in forward
    input = module(input)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/DenseFusion-Pytorch-1.0/lib/extractors.py", line 32, in forward
    out = self.conv1(x)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/module.py", line 547, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/conv.py", line 343, in forward
    return self.conv2d_forward(input, self.weight)
  File "/home/xsy/anaconda3/envs/pytorch1.0/lib/python3.5/site-packages/torch/nn/modules/conv.py", line 340, in conv2d_forward
    self.padding, self.dilation, self.groups)
RuntimeError: cuDNN error: CUDNN_STATUS_EXECUTION_FAILED

加入代码：

torch.backends.cudnn.enabled = False

报错：RuntimeError: CUDA error: CUBLAS_STATUS_EXECUTION_FAILED when calling `cublasSgemm( handle, opa, opb, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc)`