fraunhofer-iis / prosper_nn Goto Github PK

There are unregistered parameters in the model torch modules. If the model is moved to gpu these parameters stay at cpu.

The following paper presents three GRU variants. Variant 3 ist similar to the LSTM implementation in ECNN and HCNN. One difference is that r_t is fixed to a vector with ones in our implementation.

R. Dey and F. M. Salem, "Gate-variants of Gated Recurrent Unit (GRU) neural networks,"
2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS),
Boston, MA, USA, 2017, pp. 1597-1600, doi: 10.1109/MWSCAS.2017.8053243

Matplotlib dependency is 3.7 not 4.0

For HCNNCell and HcnnLstmCell the teacher_forcing output is scaled in the dropout layer if teacher_forcing < 1. Therefore, the correction of the state is wrong and the intermediate state does not contain the values of the observation.