LSTM - remove redundant bias

karpathy / char-rnn

Multi-layer Recurrent Neural Networks (LSTM, GRU, RNN) for character-level language models in Torch

11.53k stars 2.58k forks source link

Open pgmmpk opened 7 years ago

pgmmpk commented 7 years ago

Since i2h and h2h are later added, no need to create bias parameter in each linear transform.

This is not an error, just a waste of few parameters.