Closed mz24cn closed 7 years ago
The code implements C++ version charRNN for mxnet\example\rnn\char-rnn.ipynb with MXNet.cpp API. The generated params file is compatiable with http://data.dmlc.ml/mxnet/data/char_lstm.zip. train() and predict() has been tested (compiler: VS2015).
Sorry @mz24cn, after I fixed travis, I found there are lint errors in your code. Please resubmit this pull request and handle with the lint errors.
The code implements C++ version charRNN for mxnet\example\rnn\char-rnn.ipynb with MXNet.cpp API. The generated params file is compatiable with http://data.dmlc.ml/mxnet/data/char_lstm.zip. train() and predict() has been tested (compiler: VS2015).