Kartik Ramesh

Results 4 comments of Kartik Ramesh

@kaavee315 I think this issue can be closed thanks to #1453

Thank you for the response! For now I will use `benchmark_serving.py` to compare. Is supporting data parallel in `benchmark_throughput.py` planned for the future? I can help contribute if so.

I am also facing this issue. Is there any workaround for this?