Prerequisites
- Before you begin, complete the prerequisites and install the required operators and components as described in the Caikit-TGIS-Serving readme file.
Procedure
-
Navigate to the
/demo/kserve/
directory. -
Enable and configure User Workload Monitoring.
oc apply -f custom-manifests/metrics/uwm-cm-enable.yaml oc apply -f custom-manifests/metrics/uwm-cm-conf.yaml
Optionally, you can change the retention time by editing the
uwm-cm-conf.yaml
file. -
Deploy a model by using either of these options:
-
By following step-by-step commands as described in Deploying an LLM model with the Caikit+TGIS Serving runtime.
-
By running scripts as described in Using scripts to deploy an LLM model with the Caikit+TGIS Serving runtime.
-
Access Caikit, TGIS and Istio metrics:
a. From the Openshift Console, select Observe --> Metrics.
b. Search for any
caikit_*
,tgi_*
oristio_*
metrics.