Average pooling not exported in cuda_convnet pooling code

[Vaishaal]

Would be useful to have.

[nouiz]

pylearn2 don't have developer anymore. There is other framework that are
developed on top of Theano like Blocks, Keras and Lasagne.

In Theano, we now don't need to use cuda_convnet for pooling, so you could
update the code that call cuda_convnet pooling and make it call
theano.tensor.signal.pool.pool_2d() instead and pass the parameter average.

On Thu, Jul 21, 2016 at 1:37 PM, Vaishaal [email protected] wrote:

Would be useful to have.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#1574, or mute the thread
https://github.com/notifications/unsubscribe-auth/AALC-4chyIUt5hS6ponw9qAzQtflWv4yks5qX65jgaJpZM4JSCdv
.