Merge remote-tracking branch 'upstream/main'

Author: dchourasia

OWNERS

@@ -6,6 +6,7 @@ approvers:
   - jiripetrlik
   - kpostoffice
   - sutaakar
+  - szaher
   - varshaprasad96
 reviewers:
   - abhijeet-dhumal
@@ -17,6 +18,7 @@ reviewers:
   - jiripetrlik
   - kpostoffice
   - sutaakar
+  - szaher
   - varshaprasad96
 emeritus_approvers:
   - anishasthana # 2024-09-04

go.mod

@@ -7,7 +7,7 @@ toolchain go1.21.5
 require (
 	github.com/kubeflow/training-operator v1.7.0
 	github.com/onsi/gomega v1.31.1
-	github.com/project-codeflare/codeflare-common v0.0.0-20241015133940-3e0d9b3a23ad
+	github.com/project-codeflare/codeflare-common v0.0.0-20241108084652-0d76fd215a22
 	github.com/prometheus/client_golang v1.20.4
 	github.com/prometheus/common v0.57.0
 	github.com/ray-project/kuberay/ray-operator v1.1.0-alpha.0

go.sum

@@ -365,8 +365,8 @@ github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZb
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/project-codeflare/appwrapper v0.8.0 h1:vWHNtXUtHutN2EzYb6rryLdESnb8iDXsCokXOuNYXvg=
 github.com/project-codeflare/appwrapper v0.8.0/go.mod h1:FMQ2lI3fz6LakUVXgN1FTdpsc3BBkNIZZgtMmM9J5UM=
-github.com/project-codeflare/codeflare-common v0.0.0-20241015133940-3e0d9b3a23ad h1:rZEFsEa4VQXw/U2AHt2HJ4+e55CXYElPIhOSr3+3e9o=
-github.com/project-codeflare/codeflare-common v0.0.0-20241015133940-3e0d9b3a23ad/go.mod h1:v7XKwaDoCspsHQlWJNarO7gOpR+iumSS+c1bWs3kJOI=
+github.com/project-codeflare/codeflare-common v0.0.0-20241108084652-0d76fd215a22 h1:wzIJHoGAmNZupO3ZI7gbONuXgIUireabHsZvMt+3fqQ=
+github.com/project-codeflare/codeflare-common v0.0.0-20241108084652-0d76fd215a22/go.mod h1:v7XKwaDoCspsHQlWJNarO7gOpR+iumSS+c1bWs3kJOI=
 github.com/prometheus/client_golang v0.9.1/go.mod h1:7SWBe2y4D6OKWSNQJUaRYU/AaXPKyh/dDVn+NZz0KFw=
 github.com/prometheus/client_golang v1.0.0/go.mod h1:db9x61etRT2tGnBNRi70OPL5FsnadC4Ky3P0J6CfImo=
 github.com/prometheus/client_golang v1.7.1/go.mod h1:PY5Wy2awLA44sXw4AOSfFBetzPP4j5+D6mVACh+pe2M=

instructlab/standalone/standalone.py

@@ -50,8 +50,8 @@
 logger = logging.getLogger(__name__)
 
 # IMAGES
-DS_IMAGE = "quay.io/opendatahub/workbench-images:jupyter-datascience-ubi9-python-3.11-20241004-609ffb8"  # pylint: disable=line-too-long
-RHELAI_IMAGE = "registry.redhat.io/rhelai1/instructlab-nvidia-rhel9:1.2"
+DS_IMAGE = "quay.io/opendatahub/workbench-images@sha256:7f26f5f2bec4184af15acd95f29b3450526c5c28c386b6cb694fbe82d71d0b41"  # pylint: disable=line-too-long
+RHELAI_IMAGE = "registry.redhat.io/rhelai1/instructlab-nvidia-rhel9@sha256:b3dc9af0244aa6b84e6c3ef53e714a316daaefaae67e28de397cd71ee4b2ac7e"  # pylint: disable=line-too-long
 
 # SDG
 DEFAULT_REPO_URL = "https://github.com/instructlab/taxonomy.git"

tests/ilab/README.md

@@ -0,0 +1,45 @@
+# InstructLab Standalone Tool Integration Tests
+
+### Prerequisites
+
+* Admin access to an OpenShift cluster
+
+* Installed OpenDataHub or RHOAI, enabled all Distributed Workload and Serving components
+
+* OpenShift contains StorageClass with ReadWriteMany access mode
+
+* OpenAI compliant Judge model deployed and served on an endpoint reachable from OpenShift
+
+* Installed Go 1.21
+
+## Required environment variables
+
+### Environment variables to download SDG and upload trained model
+
+* `AWS_DEFAULT_ENDPOINT` - Storage bucket default endpoint
+* `AWS_ACCESS_KEY_ID` - Storage bucket access key
+* `AWS_SECRET_ACCESS_KEY` - Storage bucket secret key
+* `AWS_STORAGE_BUCKET` - Storage bucket name
+* `SDG_OBJECT_STORE_DATA_KEY` - Path in the storage bucket where SDG bundle is located
+
+### Environment variables for connection to Judge model
+
+* `JUDGE_ENDPOINT` - Endpoint where the Judge model is deployed to (it should end with `/v1`)
+* `JUDGE_NAME` - Name of the Judge model
+* `JUDGE_API_KEY` - API key needed to access the Judge model
+* `JUDGE_CA_CERT_FROM_OPENSHIFT` (Optional) - If Judge model is deployed in the same OpenShift instance and the OpenShift certificate is insecure then set this env variable to `true`. It will indicate to the test to set OpenShift CA certificate as trusted certificate.
+
+### Misc environment variables
+
+* `TEST_NAMESPACE` (Optional) - Specify test namespace which should be used to run the tests
+* `TEST_ILAB_STORAGE_CLASS_NAME` (Optional) - Specify name of StorageClass which supports ReadWriteMany access mode. If not specified then test assumes StorageClass `nfs-csi` to exist.
+* `RHELAI_WORKBENCH_IMAGE` (Optional) - Specify Workbench image to be used to run Standalone tool. If not specified then test uses Workbench image `quay.io/opendatahub/workbench-images:jupyter-datascience-ubi9-python-3.11-20241004-609ffb8`.
+Provided image should contain `click==8.1.7` and `kubernetes==26.1.0` packages.
+
+## Running Tests
+
+Execute tests like standard Go unit tests.
+
+```bash
+go test -run TestInstructlabTrainingOnRhoai -v -timeout 180m ./tests/ilab/
+```