Closed mtshikomba closed 4 years ago
I used the code below when doing training and evaluation, if you check your train.py script you'll find the line device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
just replace it with this
device = torch.device("cpu")
.
For some reason i could only do this on a CUDA enabled machine, meaning that i could not run this project on a non NVIDIA Graphics card machine. Not sure If am explaining this clearly.
@TangeniThePyGuru does it work with the code, It seems like the same, if I do not install cuda, the code above is just the same as code below. here is the error, CUDA driver library cannot be found. If you are sure that a CUDA driver is installed, try setting environment variable NUMBA_CUDA_DRIVER with the file path of the CUDA driver shared library. :
Do you have a idea to solve that?
have u solved the problem to run pointpillars or second on CPU? THANKS