Multi-GPU on windows

I can use simple PyTorch statements on Windows 10 and they work OK. But fancier applications, such as OpenNMT require some muti-procressing functions that are available only in Linux (see trace below). Is there any plan to port these to Windows in the near future?

c:\work\OpenNMT-py>python train.py -data data/demo -save_model demo-model -world_size 2 -gpu_ranks 0 1
Traceback (most recent call last):
File “train.py”, line 217, in
main(opt)
File “train.py”, line 93, in main
raise Exception(msg)
Exception:

– Tracebacks above this line can probably
be ignored –

Traceback (most recent call last):
File “c:\work\OpenNMT-py\train.py”, line 148, in run
gpu_rank = onmt.utils.distributed.multi_init(opt, device_id)
File “c:\work\OpenNMT-py\onmt\utils\distributed.py”, line 25, in multi_init
torch.distributed.init_process_group(
AttributeError: module ‘torch.distributed’ has no attribute ‘init_process_group’