I built a Bayesian neural network and the loss value is smaller than 0, but the prediction result is reasonable. Is this normal?
Hi Haoyu,
the loss is just a function that is getting minimized during training. There is notthing special in the value 0.
Typically, you can define it in a way that makes 0 mean “no error”, “perfect result” or a similar concept but it’s not necessary for the training to converge.
Thank you, the results are indeed reasonable.