I wrote a bunch of code on github related to model inference. The inner loops are executed trillions of times in regular usage. Part of the value my code prov

I wrote a bunch of code on github related to model inference. The inner loops are executed trillions of times in regular usage. Part of the value my code provides is I've spent several hundred hours going through it with vtune on different xeons. I regularly get mail saying my code is structured incorrectly or critiquing the code as if I'm a novice developer. It's very frustrating.