Skip to content

[model-gateway] add embedding tests#16583

Merged
slin1237 merged 1 commit intomainfrom
smg-ci-n/11
Jan 6, 2026
Merged

[model-gateway] add embedding tests#16583
slin1237 merged 1 commit intomainfrom
smg-ci-n/11

Conversation

@slin1237
Copy link
Copy Markdown
Collaborator

@slin1237 slin1237 commented Jan 6, 2026

  • Update conftest.py to properly detect e2e tests without explicit model markers and include DEFAULT_MODEL in pool requirements
  • Add embedding test suite (test_basic.py, test_correctness.py)
  • Add health check caching to avoid redundant worker health checks

Checklist

Review Process

  1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
  2. Get approvals from CODEOWNERS and other reviewers.
  3. Trigger CI tests with comments (/tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci) or contact authorized users to do so.
  4. After green CI and required approvals, ask Merge Oncalls to merge.

@gemini-code-assist
Copy link
Copy Markdown
Contributor

Summary of Changes

Hello @slin1237, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly expands the testing coverage for the model gateway by introducing dedicated end-to-end tests for the embedding API. It ensures that embedding generation is functional, numerically correct, and consistently handled by the router across different backend configurations. The changes also include crucial updates to the test infrastructure to better manage model launching and provide clearer debugging information, ultimately enhancing the stability and maintainability of the embedding service.

Highlights

  • New Embedding Test Suites: Introduced comprehensive end-to-end test suites for the embedding API, covering basic functionality, semantic correctness against HuggingFace references, and router-specific behaviors. These tests ensure the reliability and accuracy of embedding generation across gRPC and HTTP backends.
  • Enhanced E2E Test Configuration: Updated conftest.py to intelligently detect e2e tests that lack explicit model markers and automatically include the DEFAULT_MODEL in the pool requirements. This improves test setup robustness and reduces boilerplate.
  • Embedding Model Worker Support: Modified the model pool infrastructure to correctly identify and launch embedding models by adding an --is-embedding flag to worker commands based on model specifications. This ensures embedding-specific workers are properly configured.
  • Improved Logging for Debugging: Added more detailed logging for gateway commands and model health checks, including the full command string for gateway launches and both router and worker URLs during health checks, which will aid in debugging.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Ignored Files
  • Ignored by pattern: .github/workflows/** (2)
    • .github/workflows/pr-test-pd-router.yml
    • .github/workflows/pr-test-rust.yml
Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature Command Description
Code Review /gemini review Performs a code review for the current pull request in its current state.
Pull Request Summary /gemini summary Provides a summary of the current pull request in its current state.
Comment @gemini-code-assist Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help /gemini help Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

  1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution.

Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request introduces a comprehensive test suite for the embedding functionality, covering basic API usage, correctness against a reference implementation, and router behavior. The modifications to conftest.py to intelligently manage a default model for tests are a solid improvement. While the new tests are well-structured, I've identified a key area for improvement: several test helpers make individual API calls in a loop instead of sending a single batch request. This approach is inefficient and misses the opportunity to properly test the system's batching capabilities. My review includes suggestions to refactor these helpers to use batch API calls, which will make the tests more robust and efficient. Additionally, I've pointed out a few assertions that could be strengthened for better test accuracy. Overall, this is a valuable contribution that significantly enhances test coverage.

Comment thread sgl-model-gateway/e2e_test/embeddings/test_correctness.py
Comment thread sgl-model-gateway/e2e_test/router/test_embeddings.py Outdated
Comment thread sgl-model-gateway/e2e_test/embeddings/test_basic.py
Comment thread sgl-model-gateway/e2e_test/embeddings/test_basic.py
@slin1237 slin1237 added the run-ci label Jan 6, 2026
@slin1237 slin1237 force-pushed the smg-ci-n/11 branch 6 times, most recently from 49e21ff to 91fc6d1 Compare January 6, 2026 20:34
- Update conftest.py to properly detect e2e tests without explicit model
  markers and include DEFAULT_MODEL in pool requirements
- Add embedding test suite (test_basic.py, test_correctness.py)
- Add health check caching to avoid redundant worker health checks
@slin1237 slin1237 merged commit a49b9a6 into main Jan 6, 2026
59 of 62 checks passed
@slin1237 slin1237 deleted the smg-ci-n/11 branch January 6, 2026 21:24
michaelzhang-ai pushed a commit to michaelzhang-ai/sglang that referenced this pull request Jan 7, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants