Speaking about random sampling during inference, most ML models are doing it rather inefficiently. Here’s a better way: https://github.com/Const-me/Cgml/blob/m

Speaking about random sampling during inference, most ML models are doing it rather inefficiently. Here’s a better way: https://github.com/Const-me/Cgml/blob/master/Readme.md#random-sampling-shaders https://github.com/Const-me/Cgml/blob/master/Readme.md#rando... My HLSL is easily portable to CUDA, which has `__syncthreads` and `atomicInc` intrinsics.