I am getting error in the music inference part and predict and sample part of the assignment Improvise_a_Jazz_Solo_with_an_LSTM_Network_v4

Course Q&A Deep Learning Specialization Sequence Models
1

May some one figure out? I am also attaching the images here include the code and error.

2

First, please do not paste your code.

For your error, please double check the guidance in a Jupyter notebook.

Get the index of the maximum value of the predicted output using tf.math.argmax along the last axis.

3

Dear Sir,
May you please tell me " Get the index of the maximum value of the predicted output using tf.math.argmax along the last axis.", how to do this?

4

If you look at tf.math.argmax,
you see one parameter, “axis”. What we want to set is the last axis. Python has a good way to do that. That’s “-1”. Actually, “-1” is used at the several cases and useful. (axis=0 is the first axis.)

5

Dear Sir,
I think that axis=1 is for last axis.

6

It depends on the test data. It may work for this particular test case. But, in some cases, axis=1 may not be “the last” one if the dimension of data is more than 3. To make our logic woks generally, setting “-1” is much safer than setting a fixed axis.

7

Thanks a lot, sir for your help.

8

Since the grader always uses a different data set than the notebook, it’s not a safe assumption that the number of axes are the same.