Week 1 Assignment: Dinosaurus_Island_Character_level_language_model

Hi everyone,
Can someone please help me understand the intuition behind this model and how it works?
There is no proper intuition provided in the exercise.
As we are using y[t] = x[t+1]. How it trains and predicts efficiently to generate new names?


Hello @Muhammad-Kalim-Ullah,

The process that is described by the image you have shared is about “sampling some sequences” from a trained model. Ofcourse, the model has not yet been trained by the time you work on that part of the assignment, but you are just implementing the sampling process so that you can use it later.

I strongly suggest you to watch this video (C5 W1 Sampling Novel Sequences) because in which Andrew has very clearly explained the purpose of doing this sampling process and how to carry it out. The text description in that part of the assignment is basically the more detailed version of what Andrew talked about in the video, so I believe the text there is going to be the best supplementary material for you to complete the assignment.

At the end of the video, you will also see what kind of things you can expect from such sampling process. Since the model in the assignment is specifically for giving names, hopefully, after you trained the model in the later part of the assignment, you will be able to see some good names.


1 Like