modelmesh-serving
modelmesh-serving copied to clipboard
kserve
Controller for ModelMesh
Currently, all the ModelMesh repositories use UBI version 8 as base images. We should upgrade to UBI version 9. FYI @njhill @tjohnson31415
**Describe the bug** While following the tutorial '[Creating a custom serving runtime in KServe ModelMesh](https://developer.ibm.com/tutorials/awb-creating-custom-runtimes-in-modelmesh/)' from the IBM site, I was trying to make a small adjustment, loading the **sklearn...
Added s390x in node affinity rule as already work is going on to support s390x for tensorflow in runtime-adaptor
#### Motivation Doc updates post v0.12.0 release. #### Modifications Update the following files in the `main` branch, replacing all occurrences pointing to the old `release-*` branch or the previous release...
Release process reference: https://github.com/kserve/modelmesh-serving/blob/3e755a909cbdcbadb6fb86cbd48c83ae46e68f9a/docs/release-process.md # Release Process ## Prepare the Release Check pending vulnerability fixes, features and bugs: - [x] [`modelmesh`](https://github.com/kserve/modelmesh/security) - https://github.com/kserve/modelmesh/pull/136 - [x] [`modelmesh-minio-examples`](https://github.com/kserve/modelmesh-minio-examples/security) - [x] [`modelmesh-runtime-adapter`](https://github.com/kserve/modelmesh-runtime-adapter/security) -...
### Description I am working with ModelMesh Serving deployed on a Kubernetes cluster and I am looking for a way to control the number of replicas for a specific model....
We came across a vulnerability where the `controller-runtime` pulls, as part of the `[email protected]`, a dependency that has the following high vulnerability: - https://www.cve.org/CVERecord?id=CVE-2023-37788 As we can see in the...
I don't see any information in the documentation about how Predictors in ModelMesh are autoscaled... How is the autoscaling of (the "copies") a Predictor running in ModelMesh configured/managed? Which component...
**Describe the bug** If you don't have `ginkgo` installed and try to run the `fvt` tests it will fail. ```bash make fvt ginkgo -v -procs=2 --fail-fast fvt/predictor fvt/scaleToZero fvt/storage fvt/hpa...
Applies to custom runtimes which use gRPC service interfaces other than the standard KFSv2 one. Currently the `mm-vmodel-id` header must still be used for these. Add a section in the...
