Hi,
I was wondering to make my neural network more robust.
I tried to change the model after the feedforward stage but then I get Runtime errors.
What is the best way to inject faults to a model during training?
Hi,
I was wondering to make my neural network more robust.
I tried to change the model after the feedforward stage but then I get Runtime errors.
What is the best way to inject faults to a model during training?
someone who can help me?