karpathy / char-rnn

Multi-layer Recurrent Neural Networks (LSTM, GRU, RNN) for character-level language models in Torch
11.59k stars 2.58k forks source link

usual prediciton error #167

Closed okvol closed 8 years ago

okvol commented 8 years ago

I'm new to this area and my plain RNN (not using this code) has next-character prediction error ~55% on the Shakespeare data. Is it good or bad? What prediction errors should I expect with plain RNN and LSTM, respectively? Some rough ranges would be very appreciated. Thanks!