Suboptimal convergence when compared with TensorFlow model

vision
18

I thought I was the only one! Same problem here: RNN and Adam: slower convergence than Keras

When I’ll have time I’ll try with other optimizers.

EDIT: same situation with RMSProp.