4 ms·

yes, and i bet they didn't use 2500 CPUs either

• jsty 7y ago
  Given that 2500 n1-standard-1 preemptible instances in us-central-1 would cost a grand total of $25/hr, it's certainly not a 'must be Google or taller' requirement
• pplonski86 7y ago
  The first thing is, the Neural Networks are not golden ML algorithm that is working always the best. I'm not surprised to see GBDT methods working better than NN. The NNs are very powerful because they can accept the wide-range of data types (tabular, images, voice) which is not possible with GBDT (maybe not yet). The second, I think architecture search fo NNs is very inefficient right now. Most of the methods are training new NN in every attempt. There is a lack for method that can start with any NN architecture and grow/prune it dynamically. Take a look at this poster, with dynamic topology adjustemnt for MLP http://www.ire.pw.edu.pl/~rsulej/NetMaker/icaisc/icaisc_poster.pdf http://www.ire.pw.edu.pl/~rsulej/NetMaker/icaisc/icaisc_post...
  • mark_l_watson 7y ago
    You might like AdaNet that doe’s architecture search in one TensorFlow session and seems efficient (at least my use of it).