enable multi-gpu training with python interface ?

[zeakey]

Hey there,

I'm now trying out a training procedure with complicated internal operations, I plan to do it with caffe's python interface.

But I found it seems that caffe doesn't support multi-gpu training with python interface, am I right ?

Is it painful to implement such a multi-gpu support for python-interface ?

[trachelr]

+1 any insight about this issue??

[willyd]

You can have a look at this PR #4563

[cypof]

It should work now, #4563 was merged.

[trachelr]

Thanks ! Le 17 janv. 2017 21:44, "Cyprien Noel" <notifications@github.com> a écrit :

…

Closed #4119 <#4119>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#4119 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEDleg0xOttHUNb7m5JCHZvOEaJnU-8gks5rTShAgaJpZM4IayV4> .

[soulslicer]

Is there any example code that shows you how to do this