What does fallback_function actually meaning when torch.autograd.profiler.profile called

HI all:
when I compare the inference time cost between libtorch(c++) and torchscript(python),with LaneGCN network,I found that c++ inference time more slow than python, I have no idea about that.
actually, when compare torchscript with the original ckpt model, torchscript inference sometimes run more slower than ckpt inference, when I use torch.autograd.profiler.profile to debug these slower time point, I got this pictual below:


I search a lot, but don’t know what the fallback_function mean, it cost a lot, does it mean the gpu memory realloc or something else?