gpu=1, but code can run in 2 gpus.

[WindSmileValley]

Hi, I meet one strange question.
case 1:
if sei_trainer = Trainer(
gpus=2,
strategy='dp')
and
self.model = torch.nn.DataParallel(self.model,device_ids = [0,1]) # which is in a LightningModule class
there is question " RuntimeError: Expected tensor for argument #1 'input' to have the same device as tensor for argument #2 'weight'; but device 1 does not equal 0 (while checking arguments for cudnn_convolution)

but
case 2:
if sei_trainer = Trainer(
gpus=1,
strategy='dp')
and
self.model = torch.nn.DataParallel(self.model,device_ids = [0,1]) # which is in a LightningModule class
the code can run in 2 gpus.

I can not understant why this is happen and how to fix it.

Thank you all.