Playing around with sherjilozair's char-rnn-tensorflow using Japanese language datasets.
Currently working with a copy of こころ by 夏目漱石 available here: http://www.aozora.gr.jp/cards/000148/files/773_14560.html as well as a business Japanese dataset originally in the Conll format.