You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@submarine.apache.org by "Byron Hsu (Jira)" <ji...@apache.org> on 2021/06/21 10:05:00 UTC
[jira] [Updated] (SUBMARINE-858) Implement log_param and log_metric
for distributed scenarios
[ https://issues.apache.org/jira/browse/SUBMARINE-858?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Byron Hsu updated SUBMARINE-858:
--------------------------------
Description:
1. Calling start(), ModelsClient will parse the worker_id + experiment_id and creates a Mlflow run.
2. The Mlflow run name is worker_id, and the Mflow experiment name is experiment_id.
3. Calling log_param() or log_metric(), their info will be directed to the Mlflow runs previously set.
was:
1. ModelsClient parses the worker_id + experiment_id and creates a Mlflow run.
2. The Mlflow run name is worker_id, and the Mflow experiment name is experiment_id.
> Implement log_param and log_metric for distributed scenarios
> ------------------------------------------------------------
>
> Key: SUBMARINE-858
> URL: https://issues.apache.org/jira/browse/SUBMARINE-858
> Project: Apache Submarine
> Issue Type: Sub-task
> Components: model management
> Reporter: Byron Hsu
> Priority: Major
>
> 1. Calling start(), ModelsClient will parse the worker_id + experiment_id and creates a Mlflow run.
> 2. The Mlflow run name is worker_id, and the Mflow experiment name is experiment_id.
> 3. Calling log_param() or log_metric(), their info will be directed to the Mlflow runs previously set.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@submarine.apache.org
For additional commands, e-mail: dev-help@submarine.apache.org