3 ms·

First time I heard of DRAGNN, my mind is blown: https://github.com/tensorflow/models/blob/master/research/syntaxnet/g3doc/DRAGNN.md https://github.com/tensorflo

• igravious 9y ago
  Posted here: https://news.ycombinator.com/item?id=13913962 https://news.ycombinator.com/item?id=13913962 8 months ago by @Katydid but got zero traction. I think we're perhaps at peak NN :/
• syllogism 9y ago
  Mostly DRAGNN is a lot of hard work to compensate for how difficult Tensorflow makes your life if you're trying to write this type of model. If you use a library like Chainer, PyTorch or DyNet, the problem this solves simply never occurs.
  • Eridrus 9y ago
    This may be true for training, but I have not heard of a real deployment story for low latency NLP models in any of these define by run frameworks, which this attempts to solve.
    • syllogism 9y ago
      I haven't benchmarked, but I think you'll find DyNet to be significantly faster than DRAGNN in that setting. I'm also not sure DRAGNN is a good direction to head in, if that's the problem. You'll never ever be able to change anything in it, so if performance isn't good you're stuck. Writing the forward pass isn't very difficult, so I'd much rather be able to replace parts of a network with optimised code if necessary.
      • Eridrus 9y ago
        Would be good to see some non-batch inference benchmarks for this. I couldn't find any numbers online.