Back to archive

Thread

3 tweets

2
@Al_Grigor @jruby In retrospect it is interesting how jblas tried to solve a problem once (fast linear algebra on the JVM), but then tensorflow and friends took it to the next level by putting it on the GPU, and adding differentiation.
3
@Al_Grigor @jruby DeepLearning4j was the one that did that on the JVM. But AFAIK nobody picked up on JRuby...