C2_W1_Lab02_CoffeeRoasting_TF Counting of weights and tensorflow

week2 , Coursera | Online Courses & Credentials From Top Educators. Join for Free | Coursera can someone help in how to approach this lab i am stuck at this lab am not able to go ahead since i am not understanding the methods that are called since am new to TF, i was able to understand until the point of Normalization, then i lost the track when tile was called how does it help the training and converge the training also how did we reach at the number of parameters vector X has 2 features and we have defined the neuron function yet since we have 3 neurons in layer 1 how do we get9 parameters and all the below code went above my head

For the coffee roasting lab, you just have to read the markup and run all of the cells in sequence.

The materials should also be covered in the lectures.

thanks for the response, wanted to ask the meaning behind training for multiple epochs and the reason behind it and also how does tile function help here by appending the same data to the training set, i am just beginner started learning now in other cases like linear or logistic regression we fix the iterations and the weights and bias get converged but in case of nueral networks why are we repeating the fit for say 2000 data points and repeat the training for another 2000 points whats changing here for each epoch