Merge branch 'master' into fix/macos-mktemp

ods_ci/tasks/Resources/RHODS_OLM/install/oc_install.robot

@@ -17,7 +17,7 @@ ${SERVICEMESH_OP_NAME}=     servicemeshoperator
 ${SERVICEMESH_SUB_NAME}=    servicemeshoperator
 ${AUTHORINO_OP_NAME}=     authorino-operator
 ${AUTHORINO_SUB_NAME}=    authorino-operator
-${AUTHORINO_CHANNEL_NAME}=  managed-services
+${AUTHORINO_CHANNEL_NAME}=  tech-preview-v1
 ${RHODS_CSV_DISPLAY}=    Red Hat OpenShift AI
 ${ODH_CSV_DISPLAY}=    Open Data Hub Operator
 ${CUSTOM_MANIFESTS}=    ${EMPTY}

ods_ci/test-variables.yml.example

@@ -82,3 +82,5 @@ APPLICATIONS_NAMESPACE: redhat-ods-applications
 MONITORING_NAMESPACE: redhat-ods-monitoring
 OPERATOR_NAMESPACE: redhat-ods-operator
 NOTEBOOKS_NAMESPACE: rhods-notebooks
+PIP_INDEX_URL: https://pypi.org/simple
+PIP_TRUSTED_HOST: pypi.org

ods_ci/tests/Resources/Files/llm/model_expected_responses.json

@@ -210,6 +210,11 @@
                     "vllm-runtime": {
                         "completions_response_text": "{\"object\":\"text_completion\",\"created\":,\"model\":\"granite-8b-code-base\",\"choices\":[{\"index\":0,\"text\":\" city in California, one of the most populous cities in the United States\",\"logprobs\":null,\"finish_reason\":\"length\",\"stop_reason\":null}],\"usage\":{\"prompt_tokens\":5,\"total_tokens\":20,\"completion_tokens\":16}}"
                     }
+                },
+                "e5-mistral-7b":{
+                    "vllm-runtime": {
+                        "embeddings_response_text": "The actual embedded output is too large, and we don't check or match the response"
+                    }
                 }
             }
         },

ods_ci/tests/Resources/Files/llm/runtime_query_formats.json

@@ -122,6 +122,18 @@
                         "response_text":    "text"
                     }
                 }
+            },
+            "embeddings"  : {
+                "http": {
+                    "endpoint": "v1/embeddings",
+                    "header": "Content-Type:application/json",
+                    "body": "{'encoding_format': 'float', 'model': '${model_name}','input': '${query_text}'}",
+                    "response_fields_map": {
+                        "response": "data",
+                        "completion_tokens":  "completion_tokens",
+                        "response_text":    "embedding"
+                    }
+                }
             }
         },
         "containers": ["kserve-container"]

ods_ci/tests/Resources/Files/llm/serving_runtimes/vllm_servingruntime_grpc.yaml

@@ -13,7 +13,7 @@ spec:
         - /models-cache
         - --port
         - "8080"
-      image: quay.io/opendatahub/vllm:stable-e392b03
+      image: quay.io/modh/vllm@sha256:60f335015eff8c99508ff421c80f5f7b23b1310d87b0d4086b6f76f9a136b5a4
       name: kserve-container
       command:
         - python3

ods_ci/tests/Resources/Files/llm/serving_runtimes/vllm_servingruntime_http.yaml

@@ -13,7 +13,7 @@ spec:
         - /models-cache
         - --port
         - "8080"
-      image: quay.io/opendatahub/vllm:stable-e392b03
+      image: quay.io/modh/vllm@sha256:60f335015eff8c99508ff421c80f5f7b23b1310d87b0d4086b6f76f9a136b5a4
       name: kserve-container
       command:
         - python3

ods_ci/tests/Resources/Files/llm/vllm/vllm_servingruntime.yaml

@@ -1,10 +1,10 @@
 apiVersion: serving.kserve.io/v1alpha1
 kind: ServingRuntime
-metadata: 
+metadata:
   name: kserve-vllm
   namespace: vllm-gpt2
-spec: 
-  annotations: 
+spec:
+  annotations:
     sidecar.istio.io/inject: "true"
     sidecar.istio.io/rewriteAppHTTPProbers: "true"
     serving.knative.openshift.io/enablePassthrough: "true"
@@ -13,46 +13,46 @@ spec:
     prometheus.io/port: '8080'
     prometheus.io/path: "/metrics/"
   multiModel: false
-  supportedModelFormats: 
+  supportedModelFormats:
     - name: vLLM
       autoSelect: true
-  containers: 
+  containers:
     - name: kserve-container
       #image: kserve/vllmserver:latest
-      image: quay.io/opendatahub/vllm:stable
-      startupProbe: 
-        httpGet: 
+      image: quay.io/modh/vllm@sha256:60f335015eff8c99508ff421c80f5f7b23b1310d87b0d4086b6f76f9a136b5a4
+      startupProbe:
+        httpGet:
           port: 8080
           path: /health
         # Allow 12 minutes to start
         failureThreshold: 24
         periodSeconds: 30
-      readinessProbe: 
-        httpGet: 
+      readinessProbe:
+        httpGet:
           port: 8080
           path: /health
         periodSeconds: 30
         timeoutSeconds: 5
-      livenessProbe: 
-        httpGet: 
+      livenessProbe:
+        httpGet:
           port: 8080
           path: /health
         periodSeconds: 100
         timeoutSeconds: 8
       terminationMessagePolicy: "FallbackToLogsOnError"
       terminationGracePeriodSeconds: 120
-      args: 
+      args:
         - --port
         - "8080"
         - --model
         - /mnt/models/gpt2
         - --served-model-name
         - "gpt2"
-      command: 
+      command:
         - python3
         - -m
         - vllm.entrypoints.openai.api_server
-      env: 
+      env:
         - name: STORAGE_URI
           value: pvc://vlmm-gpt2-claim/
         - name: HF_HUB_CACHE
@@ -62,16 +62,16 @@ spec:
         - name: NUM_GPUS
           value: "1"
         - name: CUDA_VISIBLE_DEVICES
-          value: "0"   
-      ports: 
+          value: "0"
+      ports:
         - containerPort: 8080
           protocol: TCP
-      resources: 
-        limits: 
+      resources:
+        limits:
           cpu: "4"
           memory: 8Gi
           nvidia.com/gpu: "1"
-        requests: 
+        requests:
           cpu: "1"
           memory: 4Gi
           nvidia.com/gpu: "1"

ods_ci/tests/Resources/Files/pipeline-samples/v2/pip_index_url/configmap-ds-pipeline-custom-env-vars.yaml

@@ -0,0 +1,8 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: ds-pipeline-custom-env-vars
+  namespace: my-project
+data:
+  pip_index_url: 'https://pypi.org/simple'
+  pip_trusted_host: pypi.org

ods_ci/tests/Resources/Files/pipeline-samples/v2/pip_index_url/hello_world_pip_index_url.py

@@ -0,0 +1,50 @@
+"""Hello world pipeline for pip_index_url clusters
+
+This is an example of setting pip_index_url in a pipeline task
+obtaining the value from a ConfigMap, in order to be able to run
+the pipeline in a pip_index_url environment.
+
+The pipeline reads the values from a ConfigMap (ds-pipeline-custom-env-vars)
+and creates the environment variables PIP_INDEX_URL and PIP_TRUSTED_HOST
+in the pipeline task.
+
+Note: when compiling the pipeline, the resulting yaml file only uses
+PIP_INDEX_URL (this is a limitation of kfp 2.7.0). We need to manually
+modify the yaml file to use PIP_TRUSTED_HOST.
+
+"""
+from kfp import compiler, dsl
+from kfp import kubernetes
+
+common_base_image = "registry.redhat.io/ubi8/python-39@sha256:3523b184212e1f2243e76d8094ab52b01ea3015471471290d011625e1763af61"
+
+
+@dsl.component(base_image=common_base_image,
+               pip_index_urls=['$PIP_INDEX_URL'])
+def print_message(message: str):
+    import os
+    """Prints a message"""
+    print("------------------------------------------------------------------")
+    print(message)
+    print('pip_index_url:' + os.environ['PIP_INDEX_URL'])
+    print('pip_trusted_host:' + os.environ['PIP_TRUSTED_HOST'])
+    print("------------------------------------------------------------------")
+
+
+@dsl.pipeline(name="hello-world-pipeline", description="Pipeline that prints a hello message")
+def hello_world_pipeline(message: str = "Hello world"):
+    print_message_task = print_message(message=message)
+    print_message_task.set_caching_options(False)
+
+    kubernetes.use_config_map_as_env(print_message_task,
+                                     config_map_name='ds-pipeline-custom-env-vars',
+                                     config_map_key_to_env={'pip_index_url': 'PIP_INDEX_URL'})
+
+    kubernetes.use_config_map_as_env(print_message_task,
+                                     config_map_name='ds-pipeline-custom-env-vars',
+                                     config_map_key_to_env={'pip_trusted_host': 'PIP_TRUSTED_HOST'})
+
+
+if __name__ == "__main__":
+    compiler.Compiler().compile(hello_world_pipeline,
+                                package_path=__file__.replace(".py", "_compiled.yaml"))

ods_ci/tests/Resources/Files/pipeline-samples/v2/pip_index_url/hello_world_pip_index_url_compiled.yaml

@@ -0,0 +1,85 @@
+# PIPELINE DEFINITION
+# Name: hello-world-pipeline
+# Description: Pipeline that prints a hello message
+# Inputs:
+#    message: str [Default: 'Hello world']
+components:
+  comp-print-message:
+    executorLabel: exec-print-message
+    inputDefinitions:
+      parameters:
+        message:
+          parameterType: STRING
+deploymentSpec:
+  executors:
+    exec-print-message:
+      container:
+        args:
+        - --executor_input
+        - '{{$}}'
+        - --function_to_execute
+        - print_message
+        command:
+        - sh
+        - -c
+        - "\nif ! [ -x \"$(command -v pip)\" ]; then\n    python3 -m ensurepip ||\
+          \ python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1\
+          \ python3 -m pip install --quiet --no-warn-script-location --index-url $PIP_INDEX_URL\
+          \ --trusted-host $PIP_TRUSTED_HOST 'kfp==2.7.0' '--no-deps' 'typing-extensions>=3.7.4,<5;\
+          \ python_version<\"3.9\"' && \"$0\" \"$@\"\n"
+        - sh
+        - -ec
+        - 'program_path=$(mktemp -d)
+
+
+          printf "%s" "$0" > "$program_path/ephemeral_component.py"
+
+          _KFP_RUNTIME=true python3 -m kfp.dsl.executor_main                         --component_module_path                         "$program_path/ephemeral_component.py"                         "$@"
+
+          '
+        - "\nimport kfp\nfrom kfp import dsl\nfrom kfp.dsl import *\nfrom typing import\
+          \ *\n\ndef print_message(message: str):\n    import os\n    \"\"\"Prints\
+          \ a message\"\"\"\n    print(\"------------------------------------------------------------------\"\
+          )\n    print(message)\n    print('pip_index_url:' + os.environ['PIP_INDEX_URL'])\n\
+          \    print('pip_trusted_host:' + os.environ['PIP_TRUSTED_HOST'])\n    print(\"\
+          ------------------------------------------------------------------\")\n\n"
+        image: registry.redhat.io/ubi8/python-39@sha256:3523b184212e1f2243e76d8094ab52b01ea3015471471290d011625e1763af61
+pipelineInfo:
+  description: Pipeline that prints a hello message
+  name: hello-world-pipeline
+root:
+  dag:
+    tasks:
+      print-message:
+        cachingOptions: {}
+        componentRef:
+          name: comp-print-message
+        inputs:
+          parameters:
+            message:
+              componentInputParameter: message
+        taskInfo:
+          name: print-message
+  inputDefinitions:
+    parameters:
+      message:
+        defaultValue: Hello world
+        isOptional: true
+        parameterType: STRING
+schemaVersion: 2.1.0
+sdkVersion: kfp-2.7.0
+---
+platforms:
+  kubernetes:
+    deploymentSpec:
+      executors:
+        exec-print-message:
+          configMapAsEnv:
+          - configMapName: ds-pipeline-custom-env-vars
+            keyToEnv:
+            - configMapKey: pip_index_url
+              envVar: PIP_INDEX_URL
+          - configMapName: ds-pipeline-custom-env-vars
+            keyToEnv:
+            - configMapKey: pip_trusted_host
+              envVar: PIP_TRUSTED_HOST