Search code examples
pytorchartificial-intelligence

Artificial Neural Networks | Which loss function is used for the optimization of the model?

I have a code sample from github. Link is here: github code sample

My question is "Which loss function is used for the optimization of the model?"

Adam Optimizer or CrossEntropyLoss

What are the differences between them?

Solution

  • Loss Function determines, how good is the model. It is done by comparing the predictions and actual. CrossEntropyLoss is commonly used for classification task.

    Once Loss is identified. To train neural network model. we need to update weights of model.

    It is done by calculating the gradients using the calculated loss. Where each optimizer determines how the loss has to be moved inside model. where Adam is one of the popular optimizer. It is based on momentum and RMSprop.