Run MLPerf* with Intel®-Optimized Docker* Images
… If no proxy is needed, you can skip this step: export http proxy="your host proxy" export https proxy="your host proxy" Calibrate and dump int8 model cd /opt/workdir/code/dlrm2-99.9/pytorch-cpu-int8 bash ./run calibration.sh Note: run calibration script does not need to finish, once you see roc auc… …