How to use mix link Multi-CPU of parallel computing in chainer v2.1.0

Question

In my research . I wrote 2 layers in neural network, the bottom first layer is RNN which runs on GPU, the top second layer is CPU(the algorithm model nature is more suited to CPU), I implemented it in CPU in chainer self-defined Link.

But, the CPU layer is slow , I can't wait for deadline of my paper submit. So I want to use parallel computing of this layer.

What is the best practice and fast way to implement parallel this link?

Answer 1

First, ChainerMN (not Chainer) does not support a direct way to accelerate computation within a single layer.

I would recommend to consider the following options.

• Use GPU for the "CPU-friendly" layer as well
• Use Cython or other acceleration technologies
• Use multithreading on CPU

Thanks.