Had a question. If X1 and X2 are two inputs , and assuming that there are 5 rows. Shouldn't we have two inputs in the beginning, unless X1 and X2 are rows with 4 inputs. Also the neuron in first hidden layer is 3 , which means the width has to be 3, and height is to be 2 (or 4; if X1 and X2 are rows).
Sorry if I have got it wrong or being naive. If there is a detailed video explaining the logic by you or any of the student it would be great.
Very nice. I'm also looking forward to the example of doing the BPTT of RNN by hand.
Thank you Tom!! This spreadsheets will come handy to teach students in remote areas with out Internet.
Tom: Thank you for sharing! Very useful to secure our foundational knowledge!
Had a question. If X1 and X2 are two inputs , and assuming that there are 5 rows. Shouldn't we have two inputs in the beginning, unless X1 and X2 are rows with 4 inputs. Also the neuron in first hidden layer is 3 , which means the width has to be 3, and height is to be 2 (or 4; if X1 and X2 are rows).
Sorry if I have got it wrong or being naive. If there is a detailed video explaining the logic by you or any of the student it would be great.