Questions:

$5.$Q$3$Parameters Tuning

$5\mathrm{.}1.$For the dataset in X$,$find good parameters than can achieve a very small loss.

$5\mathrm{.}2.$For example, you could get a loss $0\mathrm{.}5037178103$

$5\mathrm{.}3.$Use the original gradient descent without early stopping for this question

$5\mathrm{.}3\mathrm{.}1.$Run the gradient descent function for various settings,

$5\mathrm{.}3\mathrm{.}2.$Use eta from this set $[0\mathrm{.}00001,0\mathrm{.}0001,0\mathrm{.}001,0\mathrm{.}01,0\mathrm{.}1,0\mathrm{.}2,0\mathrm{.}3]$

$5\mathrm{.}3\mathrm{.}3.$Use n$\_$epochs from this set $[100,500,1000,5000,10000]$