[CI] Add PR-comment-only accuracy test group for A2 nightly workflow

[zhangxinyuehfad]

What this PR does / why we need it?

Move the hardcoded accuracy test group configuration out of the A2 nightly
workflow and into a versioned JSON file under tests/e2e/models/configs/.
This makes adding, removing, or rebalancing accuracy groups a config-file
change rather than a workflow edit.

Defines two sections:

• nightly: the 4 existing accuracy groups (accuracy-group-1 through 4), run on every nightly schedule and workflow_dispatch
• pr_only: groups only triggered by explicit PR comments (pr-accuracy-group-1 with gemma-3-4b-it and internlm3-8b-instruct,no overlap with nightly groups)

Changes:

• Add generate-accuracy-matrix job that reads the JSON file with jq and outputs two matrix JSON strings (nightly_matrix, pr_only_matrix)
• single-node-accuracy-tests: replace inline matrix with fromJson(needs.generate-accuracy-matrix.outputs.nightly_matrix)
• single-node-accuracy-tests-pr-only: same replacement using pr_only_matrix
• Add model_filter input to _e2e_nightly_single_node_models.yaml and a "Compute effective model list" step so that /nightly in a PR comment runs only the specified model instead of the whole group (e.g. /nightly accuracy-group-4 Qwen3-Next-80B-A3B-Instruct)
• Pass parse-trigger.outputs.filter as model_filter to both accuracy test jobs; all existing is_run / with logic is otherwise unchanged

Does this PR introduce any user-facing change?

How was this patch tested?

• vLLM version: v0.18.0
• vLLM main: vllm-project/vllm@ed359c4

[gemini-code-assist]

Summary of Changes

Hello, 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 enhances the CI/CD pipeline by introducing a new configuration file that defines accuracy test groups for the A2 platform. It specifically adds a 'pr_only' test group, allowing for targeted accuracy testing of models like 'gemma-3-4b-it' during pull request reviews. This change aims to improve the efficiency and relevance of testing by running specific accuracy checks only when needed for PRs, rather than solely in nightly builds.

Highlights

• New Configuration File: A new JSON configuration file, accuracy_groups_a2.json, was added to define accuracy test groups for the A2 platform.
• PR-Only Accuracy Test Group: Introduced a dedicated pr_only accuracy test group within the new configuration, specifically for the A2 nightly workflow.
• Model Inclusion: The gemma-3-4b-it model was added to the new pr_only accuracy test group.

🧠 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/** (1)
  • .github/workflows/schedule_nightly_test_a2.yaml

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.

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. ↩

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

[zhangxinyuehfad]

/nightly accuracy-group-1 Qwen3-VL-8B-Instruct-W8A8

[gemini-code-assist]

Code Review

What this PR does / why we need it?

This pull request introduces a new configuration file, accuracy_groups_a2.json, within tests/e2e/models/configs/. This file defines several accuracy test groups, specifying various models (such as Qwen3, ERNIE, Llama, and Gemma) to be tested on different linux-aarch64-a2b3 OS configurations. These groups are categorized for both nightly and pr_only test runs, thereby expanding the e2e testing coverage for model accuracy on A2B3 hardware.

Does this PR introduce any user-facing change?

No, this PR only adds internal test configuration.

How was this patch tested?

This patch adds a new configuration file for e2e tests. The tests themselves would utilize this configuration, and their successful execution would validate this change. No specific new tests were added for this configuration file itself.

There is no feedback to provide on the code changes.