When I calculated initial cost = (1/2*m) * sum (prediction - actual)^2

I got 239305.74 = sum (prediction - actual)^2

and 

.00127551020408 =  (1/2*m)

multiplied together I got 305.236, which is very different from your  119652.7 

Am I doing something wrong, or did you perhaps initialize theta to something other than [0,0] ? 

Thanks

There was a mistake on the pdf: it reflected values of an old version. Here are the corrected values: 

Cost before gradient descent:  305.236913265

Thetas:  [[ 23.44490618]
 [ -6.48710167]]
Cost after gradient descent:  9.33830885019

Apologies for any inconvenience.