Add baseline configuration for performance comparison of llm-d#8
Open
oshfeder wants to merge 1 commit intollm-d-incubation:mainfrom
Open
Add baseline configuration for performance comparison of llm-d#8oshfeder wants to merge 1 commit intollm-d-incubation:mainfrom
oshfeder wants to merge 1 commit intollm-d-incubation:mainfrom
Conversation
Signed-off-by: Oshrit Feder <oshritf@il.ibm.com>
rachelt44
requested changes
Apr 16, 2026
| kubectl apply -f skills/compare-llm-d-configurations/llm-d-baseline-model-server-svc.yaml -n $NAMESPACE | ||
| ``` | ||
|
|
||
| Or if that file is not available locally, create it inline: |
Collaborator
There was a problem hiding this comment.
The file is always available locally, it is part of the skill. No need to repeat it here,.
|
|
||
| ### 0.5 Check for Baseline Configuration | ||
|
|
||
| If either run is labeled as "baseline" (doesnt use llm-d scheduling) or the user indicates one configuration is a baseline run, perform the following checks: |
Collaborator
There was a problem hiding this comment.
This assumes that the llm-d stack is already up and running, but we may want to deploy it using the skill, then create the service
|
|
||
| If no endpoints exist, inform the user that pods with the label `llm-d.ai/inference-serving=true` must be running for the baseline service to work. Check the running pod labels, list them to the user and suggest which label to use for the sellector (update the baseline service accordingly). | ||
|
|
||
| 4. **Set baseline base_url**: When running the benchmark for the baseline configuration (as Run A or as Run B), ensure the Run's config.yaml uses: |
Collaborator
There was a problem hiding this comment.
This part should be moved to the step that runs the benchmark
| @@ -0,0 +1,13 @@ | |||
| apiVersion: v1 | |||
Collaborator
There was a problem hiding this comment.
please move this yaml to a sub directory called scripts (or resources)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Baseline configuration:
uses a new service endpoint (instead of lllm-d scheduling gateway)
runs on same model pods