You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mxnet.apache.org by "Denis Davydenko (JIRA)" <ji...@apache.org> on 2019/01/14 21:23:00 UTC
[jira] [Commented] (MXNET-1236) MXNet Python Inference Benchmarks
and Comparison with Scala Inference Benchmark
[ https://issues.apache.org/jira/browse/MXNET-1236?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16742508#comment-16742508 ]
Denis Davydenko commented on MXNET-1236:
----------------------------------------
Let's add alarms when Python and Java/Scala metrics drift too far apart.
> MXNet Python Inference Benchmarks and Comparison with Scala Inference Benchmark
> -------------------------------------------------------------------------------
>
> Key: MXNET-1236
> URL: https://issues.apache.org/jira/browse/MXNET-1236
> Project: Apache MXNet
> Issue Type: Task
> Components: Apache MXNet Scala API
> Reporter: Piyush Ghai
> Priority: Minor
>
> The goal of this story is to capture the inference performance using a pre-trained model for MXNet with inference being performed via Python APIs.
>
> The metrics that need to be captured are Latencies : p50, p90, p99. Throughput, memory consumption (CPU/GPU).
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@mxnet.apache.org
For additional commands, e-mail: issues-help@mxnet.apache.org