I have an algorithm I want to use edward for. However, I need to change the value of a placeholder after every gradient step. I know I can do the gradient steps one at a time with inference.update()
however I don’t know how to change the value of the placeholder after I take the gradient step without running ed.VariationalInference({z: qz}, data={x: x_train, beta: qbeta})
which would ruin my parameter values. Thanks,
Cameron