Po-Han Huang (NVIDIA)

icon company NVIDIA icon location I am a deep learning software engineer at NVIDIA, focusing on performance optimizations for DL inference software.

Results 6 issues of Po-Han Huang (NVIDIA)

Remove timezone shift in submission checker for power parsing because timestamps are all in UTC now

4 comment

Remove timezone shift in submission checker for power parsing because timestamps are all in UTC now

safe-to-close

[v1.1][System detail logging][WIP] Log system details in LoadGen

2 comment

Log system details in LoadGen TODOs: - [ ] Load commands from a file - [ ] Support vendor-specific commands

inference v2.1 and backlog

[NOT TESTED YET] Support new mlperf-inference logging format

1 comment

Support new mlperf-inference logging format @nv-eric-hw Could you take a look and try it out? Thanks

Add submission directory rules for MLPerf-Inference Power submission

4 comment

Add submission directory rules for MLPerf-Inference Power submission

Rename compliance_checker_log.txt to submission_checker_log.txt for clarity

4 comment

Rename compliance_checker_log.txt to submission_checker_log.txt for clarity - The script is called `submission-checker.log` - Don't cause confusion with compliance tests

Next Meeting

Where should `implementation_id` go in results/?

Where should `implementation_id` go in results/? I can find `implementation_id` under code/ and under measurements/. Where should we put the results of the corresponding `implementation_id` in measurements?