HLS4ML is software package for machine learning inference, originally targeting FPGAs. Traditional open-source machine learning package models are translated into C++ that can be configured for one’s own use-case.
HLS4ML is being extended with oneAPI to create a high-performance inference engine for Intel x86 and future platforms. The project — and also the oneAPI implementation — is work in progress. Interested parties can use and share experiences with the package and help drive new features to be added.
Find out more about the Intel® oneAPI Base Toolkit