I want to define my own Lstm model as follows:
from keras import backend as K
from keras.callbacks import ModelCheckpoint
from keras.layers.core import Dense, Ac
You should use keras.layers.Lambda
to wrap K.*
operations as a layer instead of K.*
function directly.
# change
avg_pool = K.mean(rnn_outputs, axis = 1)
max_pool = K.max(rnn_outputs, axis = 1)
# to
avg_pool = Lambda(lambda x:K.mean(x,axis=1))(rnn_outputs)
max_pool = Lambda(lambda x:K.max(x,axis=1))(rnn_outputs)