Hot coders will always be always on fire, solve the given challenge


Hello pytorch community, I’m happy that y’all are amazing and keep rolling like fire,
when i got to know about introduction to deep learning, it was a magic , unbelievable math,
we were given assignment to build a fully connected neural network from scratch without using torch.nn module and by only using softmax and sigmoid as activators, cross entropy loss as loss function
firstly it was utter trash( The code linked above) , after a week or so i figured it out,
The challenge I’m putting up is so simple, just go ahead and solve it , hints are scattered here and there in the code with intentional errors, and the first man come up with a solution wins
The challenge is because I need to understand how people thinking perspective varies !!!
happy hunting