High priority

1 Have to find a way to not run the modelsim simulation * Done (Etienne)

2 Implemation sur carte

3 Activation function *done (Nemer)

4 Check io_stream paramter in vivado *done (Nemer) Vivado_template.py

Low priority :

En standby :

2 Vanilla RNN

Keras Parameters

return_sequences: Return all the hidden states if True, and if False the layer returns only the final hidden state (For now False for us)

stateful: If True, the hidden and cell states are not reset after a sequence, but instead state reset is called manually as needed (For now False for us) Also If True, the last state for each sample at index i in a batch will be used as initial state for the sample of index i in the following batch.

dropout: A training-time thing to reduce overfitting by randomly selecting a percentage of neurons to output zeros (False for us)

regularization: Also a method to reduce overfitting (None for us)

PArtial reconfigure

-- EtienneMarieFortin - 2018-12-04