ANN: MXNet.jl v0.0.5, LSTM char-rnn tutorial for generating random sentences

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

ANN: MXNet.jl v0.0.5, LSTM char-rnn tutorial for generating random sentences

Chiyuan Zhang
Please check out this new tutorial on training a char-level LSTM recurrent neural network (RNN) to generate random sentences from arbitrary domain.

MXNet.jl is a flexible and efficient deep learning library in Julia. Here are the main changes since the last announced version:

v0.0.5 (2015.11.14)

  • char-lstm example.
  • Network visualization via GraphViz.
  • NN-factory for common models.
  • Convenient @nd_as_jl macro to work with NDArray as Julia Arrays.
  • Refactoring: Symbol -> SymbolicNode.
  • More evaluation metrics (@vchuravy, @Andy-P)

v0.0.4 (2015.11.09)

  • ADAM optimizer (@cbecker)
  • Improved data provider API.
  • More documentation.
  • Fix a bug in array data iterator (@vchuravy)
enjoy!
-- pluskid

--
You received this message because you are subscribed to the Google Groups "julia-stats" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
For more options, visit https://groups.google.com/d/optout.