RuntimeError: CUDA error: no kernel image is available for execution on the device in the second GPU

1

Traceback (most recent call last):
File “train.py”, line 264, in
).to(device)
File “C:\Users\a.abdallah\Anaconda3\envs\ocr2\lib\site-packages\torch\nn\modules\module.py”, line 425, in to
return self._apply(convert)
File “C:\Users\a.abdallah\Anaconda3\envs\ocr2\lib\site-packages\torch\nn\modules\module.py”, line 201, in _apply
module._apply(fn)
File “C:\Users\a.abdallah\Anaconda3\envs\ocr2\lib\site-packages\torch\nn\modules\rnn.py”, line 140, in _apply
self.flatten_parameters()
File “C:\Users\a.abdallah\Anaconda3\envs\ocr2\lib\site-packages\torch\nn\modules\rnn.py”, line 129, in flatten_parameters
self.batch_first, bool(self.bidirectional))
RuntimeError: CUDA error: no kernel image is available for execution on the device

2

This error is raised if some CUDA code wasn’t compiled for the compute capability of your device.
How did you install PyTorch and which GPU are you using?

3

the error becuase i installed pytorch=1.5 and the code was wriiten in 1.4 i solved it thank

4

I’m running into similar issues while trying to use PyTorch 1.5. How did you solve it?