Compilers and Interpreters

https://d2l.ai/chapter_computational-performance/hybridize.html

Without torchscript: 1.6357 sec
With torchscript: 1.6502 sec

Regarding the above result, why torchscript improves performance?

I guess the point is it allows you to run the code in Java or other better language when you are doing CPU inference for real world application.