How can I track the progress of MLPerf Inference Datacenter runs?

[rajesh-s]

I see a bunch of "pkl" files generated in the CPU only mode and I see the following prints on my console but I am not sure where to track logs of the intermediate results:

Saving outputs to run_outputs/q18955.pkl
Samples run: 10
        BatchMaker time: 0.00016736984252929688
        Inference time: 5127.699171066284
        Postprocess time: 0.00046181678771972656
        ==== Total time: 5127.699800252914
Saving outputs to run_outputs/q7100.pkl
Samples run: 11
        BatchMaker time: 0.00014710426330566406
        Inference time: 1241.013633966446
        Postprocess time: 0.0006444454193115234
        ==== Total time: 1241.0144255161285

Run command:

cm run script --tags=run-mlperf,inference,_find-performance,_full,_r4.1 \
   --model=llama2-70b-99 \
   --implementation=reference \
   --framework=pytorch \
   --category=datacenter \
   --scenario=Offline \
   --execution_mode=test \
   --device=cpu  \
   --quiet \
   --test_query_count=50