Skip to content

Instantly share code, notes, and snippets.

@philippslang

philippslang/decl_curve_rnn_makemodel.py

Last active August 21, 2017 17:58
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save philippslang/d0dab63233acc7775119283230c175c6 to your computer and use it in GitHub Desktop.
Save philippslang/d0dab63233acc7775119283230c175c6 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
decl_curve_rnn_makemodel.py
def make_rnn(num_features, num_targets, num_timesteps, num_units):
model = kem.Sequential()
model.add(kel.LSTM(num_units, input_shape=(num_timesteps, num_features), return_sequences=True))
model.add(kel.TimeDistributed(kel.Dense(num_targets, activation='sigmoid')))
model.compile(loss='mse', optimizer='adam')
return model
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment