UPSNet icon indicating copy to clipboard operation
UPSNet copied to clipboard

DataParallel - AssertionError: Invalid device id

Open rasya-ramesh opened this issue 6 years ago • 2 comments

Traceback (most recent call last): File "upsnet/upsnet_end2end_train.py", line 407, in upsnet_train() File "upsnet/upsnet_end2end_train.py", line 176, in upsnet_train train_model = DataParallel(train_model, device_ids=[int(_) for _ in config.gpus.split(',')]).to(gpus[0]) File "/usr/local/lib/python3.6/dist-packages/torch/nn/parallel/data_parallel.py", line 133, in init _check_balance(self.device_ids) File "/usr/local/lib/python3.6/dist-packages/torch/nn/parallel/data_parallel.py", line 19, in _check_balance dev_props = [torch.cuda.get_device_properties(i) for i in device_ids] File "/usr/local/lib/python3.6/dist-packages/torch/nn/parallel/data_parallel.py", line 19, in dev_props = [torch.cuda.get_device_properties(i) for i in device_ids] File "/usr/local/lib/python3.6/dist-packages/torch/cuda/init.py", line 318, in get_device_properties raise AssertionError("Invalid device id") AssertionError: Invalid device id

The above error pops up when running upsnet_end2end_train.py. I am at a loss as to why this error is occurring.

rasya-ramesh avatar Oct 09 '19 13:10 rasya-ramesh

I wanted to add that I am using Google Colab to run this code.

rasya-ramesh avatar Oct 09 '19 13:10 rasya-ramesh

@rasya-ramesh were you able to solve it?

IAMShashankk avatar Nov 12 '20 20:11 IAMShashankk