Improvise a Jazz Solo with an LSTM Network - Input x?

Course Q&A Deep Learning Specialization Sequence Models
1

Hello All,

I am a bit confused to understand this model.

It says:

"X: This is an (m, 𝑇𝑥, 90) dimensional array.

  • You have m training examples, each of which is a snippet of 𝑇𝑥=30 musical values.
  • At each time step, the input is one of 90 different possible values, represented as a one-hot vector.
    For example, X[i,t,:] is a one-hot vector representing the value of the i-th example at time t."

But I understood that x<t+1> = y. Then then only x input would be the first one as x is fed by previous y. Right?

Is that right that X is feeding the first cell as x<1> and then x<2,3,…> will depend on it via y<> and a<>?

Thanks,
Carlos.

2

Hi Carlos,

As described a bit later “Y … is essentially the same as X, but shifted one step to the left (to the past)”.

This is going to be used to train the model, which means that the parameters determining a<> are adjusted through backpropagation based on the known values of the input x<> and the output y<>.

I hope this clarifies.