cuDNN error: CUDNN_STATUS_EXECUTION_FAILED about super-slomo

[Siemko]

You mixed both options: -sf and -fps, try only -sf or only -fps.

[Fried-Penguin-Wings]

I'll give it that a shot. Thanks.

[Fried-Penguin-Wings]

No dice. It's asking for both a variable for the slow motion factor and a variable for the target fps.

[Fried-Penguin-Wings]

I can run the command without the target fps argument, and it'll attempt it, but still results in the same error message.

[avinashpaliwal]

Can you try using pyTorch 0.4.1 and CUDA 9.2. This was the setup used for training and testing.

[CM-BF]

That is the inconsistence of your pytorch version and your CUDA version. As the author, try using pyTorch0.4.1 and CUDA 9.x or just using pyTorch1.1 and CUDA 10.1(it can work fine, too).

[Fried-Penguin-Wings]

Thanks. I'll give it a shot today.

[Fried-Penguin-Wings]

That did the trick. Thanks!