How do you increase the number of coefficients during training (as these coefficients are learnable)-For example, when you increase G from 3 to 10-, I am wondering how you are doing that (I could not understand that from your code)?
Secondly, what do you mean by "step" in your paper? does each step just process one batch of the data? (How does it relate to epoch)
How do you increase the number of coefficients during training (as these coefficients are learnable)-For example, when you increase G from 3 to 10-, I am wondering how you are doing that (I could not understand that from your code)?
Secondly, what do you mean by "step" in your paper? does each step just process one batch of the data? (How does it relate to epoch)