CODEWITHSUNDEEP
machine-learningkerasrecurrent-neural-networktf.keras
Mastiff
Mastiff

Reputation: 2240

Can I change the statefulness of RNN after training?

If I build and train my RNN based model with stateful=False, can I simply do (e.g.):

model.layers[0].stateful = True

And have it take effect as might be expected for use in prediction? I ask because in other situations (learning rate for example), the exposed parameters are "cosmetic" and don't actually affect behavior.

Upvotes: 4

Views: 216

Answers (1)

dlPFC
dlPFC

Reputation: 11

If your model is not stateful, model resets the state variable after each batch. You can create a stateful model but manually reset the state variable after each batch in prediction using callbacks (model.reset_states after on_batch_end callback).

Also, why do you want to change statefulness after training ?

Upvotes: 1

Related Questions