Caffe: 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 https://github.com/BVLC/caffe/pull/4563

[cypof]

It should work now, #4563 was merged.

[trachelr]

Thanks !

Le 17 janv. 2017 21:44, "Cyprien Noel" [e-mail] a écrit :

Closed #4119 https://github.com/BVLC/caffe/issues/4119.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/BVLC/caffe/issues/4119#event-926456096, 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