I have built a Bilstm model with attention in Tensorflow 2. However, when I am training this model, I have a vanishing gradient problem.
The code is the following:
