Dataset on Local Computer

Hello all,
I have a big dataset (for video classification) located on my local computer. Is it possible to train a PyTorch network on a remote GPU server while the dataset is located on local computer?

That would be possible, if you mount the local drive to the remote machine.
Note however that you might see a pretty big performance hit, as the data loading would go through your network and will most likely be the bottleneck of your training.