Pytorch Neural Network Model Building

Job ID: 37988138

Budget: $30 – $250 USD

I need a help session of 1-2 hours explaining the main hyper-parameters in building a neural network model, specifically:

How do I determine how many layers I need? Is it of the folly connected type or the convolutional type?

Which number of neurons should be used in each layer?

Which activation functions should be used (ReLU, Sigmoid, Tanh)?

There is a ready-made practical example (python and pytorch) on which we can experiment with different options and monitor the differences in accuracy and F1 scores.