CPU only?

[ghost]

How would I train the model with no GPU's. I am on a system with no GPU's, however, I have access to more than 16 CPU's.
Would the instructions be similar, to those for faster-rcnn?
Thank you.

"
Set USE_GPU_NMS in ./lib/fast_rcnn/config.py (or your YAML file) to False.
In ./tools/test_net.py / train_net.py, replace caffe.set_mode_gpu() with caffe_set_mode_cpu().
Plus, if you do not have a GPU, you can compile Caffe in CPU_ONLY mode (and also ignore the GPU nms code).
"

rbgirshick/py-faster-rcnn#123

[rbgirshick]

We do not plan to support CPU-only training. Some ops are currently only implemented for GPUs at this time and training will likely be too slow unless it is distributed over >>100 of CPUs. For CPU-only inference see #24.

[hxw111]

@rbgirshick You said that some ops are currently only implemented for GPUs . What about the
"AffineChannel" op? Is it implemented only for GPUs too?