Cleanup LLMEndpointType#122
Conversation
|
Skipping CI for Draft Pull Request. |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: lpiwowar The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
Important Review skippedDraft detected. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
lpiwowar
force-pushed
the
lpiwowar/providers-cleanup
branch
2 times, most recently
from
4ff11b7 to
bd5158d
Compare
This commit introduces a new Github Workflow named "Unit Tests". This workflow is responsible for running unit tests triggered by (make tests). This commit also addresses the following error that was encountered when one ran make test locally or in a job: --- unable to fetch hash for requested version: unable fetch metadata for kubebuilder-tools-1.30.0-linux-amd64.tar.gz -- got status "401 Unauthorized" from GCS --- The fixes bumps up the test dependencies (ENVTEST_VERSION release-0.18 -> release-0.22). The previous vrsion was trying to pull the the test dependencies from GCS server that has been deprecated as part of 0.19 [1]. [1] kubernetes-sigs/controller-runtime#2915
Clean up LLMEndpointType providers by dropping providers unsupported by OGX and lightspeed-core/lightspeed-stack: 1) BAM is no longer supported by IBM and was replaced by watsonx.ai [1]. 2) fake_providers is an artifact that persisted through the migration from openshift/lightspeed-operator. This provider is supported only by openshift/lightspeed-service and not lightspeed-core/lightspeed-stack [2]. Additionally, add unit tests for buildLlamaStackModels and buildLlamaStackInferenceProviders to ground the configuration. Note: this commit aims to align the providers configuration as closely as possible with the recommended configuration in lightspeed-core/lightspeed-stack [3]. It was not tested with the actual providers. Additional testing is likely required and may lead to minor tweaks in the providers code. Also, address lint issues discovered by the pre-commit checks (replace repeating strings with constants). [1] openshift/lightspeed-service#2827 [2] https://github.com/openshift/lightspeed-service/blob/main/docs/ai/providers.md [3] https://github.com/lightspeed-core/lightspeed-stack/tree/main/examples
lpiwowar
force-pushed
the
lpiwowar/providers-cleanup
branch
from
bd5158d to
1be5cd1
Compare
|
/test all |
1 participant
Clean up LLMEndpointType providers by dropping providers unsupported by OGX and lightspeed-core/lightspeed-stack:
bamis no longer supported by IBM and was replaced bywatsonx.ai[1].fake_providersis an artifact that persisted through the migration from openshift/lightspeed-operator. This provider is supported only byopenshift/lightspeed-serviceand notlightspeed-core/lightspeed-stack[2].Additionally, add unit tests for
buildLlamaStackModels()andbuildLlamaStackInferenceProviders()to ground the configuration.Note: this commit aims to align the providers configuration as closely as possible with the recommended configuration in
lightspeed-core/lightspeed-stack[3]. It was not tested with the actual providers. Additional testing is likely required and may lead to minor tweaks in the providers code.Also, address lint issues discovered by the pre-commit checks (replace repeating strings with constants).
[1] openshift/lightspeed-service#2827
[2] https://github.com/openshift/lightspeed-service/blob/main/docs/ai/providers.md
[3] https://github.com/lightspeed-core/lightspeed-stack/tree/main/examples
This PR introduces a new GitHub Workflow that runs unittests by running
make testand fixes401 Unauthorizedissue that one was encountering when runningmake test(see commit).