[AMD] CI - migrate perf test and fix stage-b-test-1-gpu-amd

[yctseng0211]

Motivation

This PR migrates the AMD CI performance and accuracy tests from a manual/imperative workflow style to using the unified test suite runner (run_suite.py).

Modifications

1. Workflow Reorganization (.github/workflows/pr-test-amd.yml)

• Renames and consolidates multiple job stages:
  • performance-test-1-gpu-part-1-amd → stage-b-test-small-1-gpu-performance-amd
  • performance-test-1-gpu-part-2-amd → stage-b-test-large-1-gpu-performance-amd
  • performance-test-1-gpu-part-3-amd → merged into other stages
  • performance-test-2-gpu-amd → stage-b-test-large-2-gpu-performance-amd
  • accuracy-test-1-gpu-amd → stage-b-test-small-1-gpu-accuracy-amd
  • accuracy-test-2-gpu-amd → stage-b-test-large-2-gpu-accuracy-amd

2. Test Registration for AMD (test/registered/perf/*.py and test/registered/eval/*.py)

• Adds register_amd_ci() in:
  • test_bench_one_batch_1gpu.py
  • test_bench_one_batch_2gpu.py
  • test_bench_serving_1gpu_large.py
  • test_bench_serving_1gpu_part1.py
  • test_bench_serving_1gpu_part2.py
  • test_bench_serving_2gpu.py
  • test_vlm_perf_5090.py
  • test_eval_accuracy_large.py
  • test_moe_eval_accuracy_large.py

3. AMD-specific Test Adjustments

• Skips the Eagle speculative decoding test on ROCm (@unittest.skipIf(is_hip(), "Skip Eagle test for ROCm"))
• Lowers the FP8 accuracy threshold for AMD (0.40 vs 0.45 for CUDA) due to different FP8 dtype (fp8_fnuz)

4. Suite Registration Updates

• Updates test/run_suite.py and scripts/ci/slash_command_handler.py to recognize the new AMD test suites

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

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 or contact authorized users to do so.
   • /tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci
4. After green CI and required approvals, ask Merge Oncalls to merge.

[gemini-code-assist]

Summary of Changes

Hello @yctseng0211, 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 integrates AMD GPU performance tests into the continuous integration pipeline. By adding specific registration calls for AMD CI, it ensures that performance benchmarks for both single and multi-GPU configurations are automatically run on AMD hardware, thereby expanding test coverage and helping to maintain performance consistency across different hardware architectures.

Highlights

• AMD CI Integration: The register_amd_ci function has been imported into the performance test files, enabling the registration of AMD-specific CI jobs.
• 1-GPU Performance Test: A new AMD CI job, stage-b-test-large-1-gpu-performance-amd, has been registered for the test_bench_one_batch_1gpu.py file, with an estimated runtime of 120 seconds.
• 2-GPU Performance Test: A new AMD CI job, stage-b-test-large-2-gpu-performance-amd, has been registered for the test_bench_one_batch_2gpu.py file, with an estimated runtime of 180 seconds.

🧠 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.

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. ↩

[gemini-code-assist]

Code Review

The pull request successfully migrates the performance tests for 1-GPU and 2-GPU configurations to include AMD CI. The changes involve importing the register_amd_ci function and then calling it with appropriate estimated times and suite names, consistent with the existing CUDA CI registrations. The modifications are clear, directly address the stated motivation, and integrate well with the existing CI registration framework.

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[yctseng0211]

These failures will be fixed by : #17432