tomrunia / PyTorchConv3D

I3D and 3D-ResNets in PyTorch
Apache License 2.0
189 stars 27 forks source link

Why don't use dataparallel on i3d? #9

Closed ziqi-zhang closed 5 years ago

ziqi-zhang commented 5 years ago

Here you disable dataparallel on i3d and I wonder the reason. Because of it using multi GPUs to train i3d becomes impossible.

I also tried to modify the code and implement multi GPUs, but I encountered "Runtime Error: Expected tensor for argument 1 'input' to have the same device as tensor for argument 2 'weight'; but device 1 does not equal 0"

punknownq commented 5 years ago

If you can understand chinese, refer to this link:https://blog.csdn.net/senius/article/details/96599955 :)

ziqi-zhang commented 5 years ago

Ok thanks!