How to use LSTM?

[dynamik1703]

Dear VinF,

thank you for your great work.

I'm using temporary the DDPG algorithm. I would like to try to use the LSTM network.

I changed in AC_net_keras the line nr. 10 for this purpose to from .NN_keras_LSTM import NN

When I'm trying to start the optimisation following error occurs: "Q_net = neural_network_critic(self._batch_size, self._input_dimensions, self._n_actions, self._random_state, True)

TypeError: init() takes 5 positional arguments but 6 were given"

How is it possible to use it?

Thanks dynamik

[VinF]

Hi dynamik, It should work thanks to #62. Can you check that everything looks fine before I merge? Thanks

[dynamik1703]

Hi VinF,

thank you very much for your quick response.

Until now it seems to work fine and I will continue checking it.

Are you planning to set the LSTM as the standard network in AC_net_keras?

Is it possible to choose it in a run_*-file the network?

Best wishes

[VinF]

For the moment the convolutional network is fine as default (I don't want to change it except if there is a reason).

You can override the default neural networks when instantiating ACnetwork in run* by providing the class of a neural network as a parameter for neural_network_critic or neural_network_actor.

Best

[dynamik1703]

Perfect. Thank you very much!