Can OpenNMT be used with OpenCL instead of CUDA ? (I do not like being dependent on NVIDIA)
There is this version of OpenCL Torch. https://github.com/hughperkins/cltorch
I cannot confirm that it works but it if you can let us know what changes would be necessary, we could look into adding them.
I was wondering if there is any progress on this topic?
did you get any results from the OpenCL library?