7 ms·
ArrayFire looks very interesting, as a novice a drop in enhancement of the nn (or nnx for dropout) package for torch would be very useful.
by psandersen 12y ago
ArrayFire looks very interesting, as a novice a drop in enhancement of the nn (or nnx for dropout) package for torch would be very useful.
- pavanky 12y agoWe are getting there. In the mean time we have an example on how you can go about building one yourself. You can see it here: https://github.com/arrayfire/arrayfire_examples/blob/master/machine_learning/neural_network.cpp https://github.com/arrayfire/arrayfire_examples/blob/master/...
- psandersen 12y agoThanks for the response, currently I'm doing R&D with easier to use high level representations and if it proves fruitful will be looking at getting the most cost effective speedups in about 6 months. Whats the license for the free edition? I would be sticking to open source code until we secure capital, but would be very interested in your commercial services down the line.