Model outputs NaN

[zhuoli7]

I have this following model:

model = Sequential()
model.add(Embedding(7551, 300, input_length = 200, mask_zero = True, weights = [embmat_300], trainable=False))
model.add(Bidirectional(LSTM(32, dropout = 0.2, recurrent_dropout = 0.2)))
model.add(Dense(2, activation = 'sigmoid'))
model.compile(loss = 'binary_crossentropy', optimizer = 'adam', metrics = ['accuracy'])

And the output of this model is something like this: Float32Array(2) [NaN, NaN] Any idea of what's wrong? Any help would be appreciated.

[zhuoli7]

Problem solved by setting gpu: true