Closed xiaoyun4 closed 4 years ago
I think that's a file from pytorch. What's your pytorch version? You can run this to check your version
python -c "import torch; print(torch.__version__)"
The pytorch version of my pc is 0.4.1.
CUDAContext.h was added in v1.0.
Update PyTorch to at least v1.2, which is the minimum version required, as it's written in the Prerequisites. I'm using the latest version (1.3.1) and it works fine
Prerequisites
- Linux or macOS
- Python 3
- NVIDIA GPU + CUDA cuDNN
PyTorch 1.2
OK,thanks a lot!
unsubscribe
在 2019-12-12 16:34:38,"Xiaoyun" notifications@github.com 写道:
Closed #5.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
unsubscribe
At 2019-12-12 11:59:36, "jajube" notifications@github.com wrote:
CUDAContext.h was added in v1.0.
Update PyTorch to at least v1.2, which is the minimum version required, as it's written in the Prerequisites. I'm using the latest version (1.3.1) and it works fine
Prerequisites Linux or macOS Python 3 NVIDIA GPU + CUDA cuDNN PyTorch 1.2
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
unsubscribe.
At 2019-12-12 11:48:10, "jajube" notifications@github.com wrote:
I think that's a file from pytorch. What's your pytorch version? You can run this to check your version
python -c "import torch; print(torch.version)"
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
unsubscribe.
At 2019-12-12 11:51:59, "Xiaoyun" notifications@github.com wrote:
The pytorch version of my pc is 0.4.1.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
the CUDA version of my pc is 9.0
When i compile the snapshot of FlowNet2 by running python scripts/download_flownet2.py, the error is "correlation_cuda.cc:4:35: fatal error: ATen/cuda/CUDAContext.h: No such file or directory"
I think the reason is that the CUDA version does not match.
Could you please tell the CUDA version?