I Have Managed To Deploy Model By Thr Clearml-Serving, Now They Are Runing On The Docker Container Engine (That Doesn'T Have Gpu In It) , What Is The Entrypoints To The Model In Order To Get Predictions?