We're going to upgrade the outlink model server Docker images, including outlink and outlink-transformer, to KServe 0.11
Description
Details
Status | Subtype | Assigned | Task | ||
---|---|---|---|---|---|
Resolved | elukey | T337213 Update to KServe 0.11 | |||
Resolved | achou | T347549 Upgrade outlink docker images to KServe 0.11 |
Event Timeline
Change 982783 had a related patch set uploaded (by AikoChou; author: AikoChou):
[machinelearning/liftwing/inference-services@main] outlink: upgrade kserve to 0.11.2
Change 982783 merged by jenkins-bot:
[machinelearning/liftwing/inference-services@main] outlink: upgrade kserve to 0.11.2
Next steps:
- Roll out the new docker images to ml-staging
- Perform load testing to ensure consistent performance
- Test the publishing of events from staging to EventGate
Regarding the last point, we are currently using the "prod" stream (mediawiki.page_outlink_topic_prediction_change.v1) for testing, which is not ideal. In T349919, there is one proposal to create a new testing stream specifically for prediction-change events. I'm leaning toward going for this option.
Change 983189 had a related patch set uploaded (by Ilias Sarantopoulos; author: Ilias Sarantopoulos):
[operations/deployment-charts@master] ml-services: update outlink topic model image on ml-staging
Change 983189 merged by jenkins-bot:
[operations/deployment-charts@master] ml-services: update outlink topic model image on ml-staging
Change 984203 had a related patch set uploaded (by AikoChou; author: AikoChou):
[operations/deployment-charts@master] ml-services: update outlink topic model image on ml-serve
Change 984203 merged by jenkins-bot:
[operations/deployment-charts@master] ml-services: update outlink topic model image on ml-serve