when i used dataparell ,i meet :\anaconda3\lib\site-packages\torch\cuda\nccl.py:16: UserWarning: PyTorch is not compiled with NCCL support warnings.warn(‘PyTorch is not compiled with NCCL support’)
But I used to use it normally ,when i update torch1.5-torch1.7,the question is coming. why?
my code is
if t.cuda.device_count() > 1:
model = nn.DataParallel(model)
if opt.use_gpu: model.cuda()
i meet the answer :Win10+PyTorch+DataParallel got warning:"PyTorch is not compiled with NCCL support"
i want to konw why torch 1.5.1 can be used dataparallel ,but 1.7.0 doesnt.
could someone answer this question for me?
I went back to 1.5.1 and couldn’t use it either
:\anaconda3\lib\site-packages\torch\cuda\nccl.py:24: UserWarning: PyTorch is not compiled with NCCL support
warnings.warn('PyTorch is not compiled with NCCL support')
0%| | 0/12097 [00:06<?, ?it/s]
Traceback (most recent call last):
File "main.py", line 188, in <module>
fire.Fire()
File "d:\anaconda3\lib\site-packages\fire\core.py", line 138, in Fire
component_trace = _Fire(component, args, parsed_flag_args, context, name)
File "d:\anaconda3\lib\site-packages\fire\core.py", line 468, in _Fire
target=component.__name__)
File "d:\anaconda3\lib\site-packages\fire\core.py", line 672, in _CallAndUpdateTrace
component = fn(*varargs, **kwargs)
File "main.py", line 103, in train
optimizer.step()
File "d:\anaconda3\lib\site-packages\torch\autograd\grad_mode.py", line 15, in decorate_context
return func(*args, **kwargs)
File "d:\anaconda3\lib\site-packages\torch\optim\adam.py", line 107, in step
denom = (exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(group['eps'])
RuntimeError: CUDA out of memory. Tried to allocate 2.18 GiB (GPU 0; 15.92 GiB total capacity; 13.74 GiB already allocated; 666.81 MiB free; 14.59 GiB reserved in total by PyTorch)
I would be very grateful
thank you
is anybody there ， i need help。 
