[CI] Migrate LoRA tests to test/registered/lora/#15176
Conversation
Summary of ChangesHello @alisonshao, 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 is a significant step in the ongoing CI reorganization effort, specifically targeting the LoRA test suite. It systematically migrates existing LoRA tests to a new, dedicated directory ( Highlights
Using Gemini Code AssistThe 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
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 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
|
|
/tag-and-rerun-ci |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request effectively migrates the LoRA tests to the new test/registered/lora/ directory as part of the ongoing CI reorganization. The changes are clean and well-executed. Key improvements include:
- Moving
lora_utils.pyto a shared location for better importability. - Migrating all specified LoRA tests and registering them with the new CI system.
- Cleaning up
sys.pathhacks in favor of proper imports, which greatly improves code quality and maintainability. - Updating the old test suites to remove the migrated tests, and adding helpful comments about the migration, especially regarding the AMD tests that need to be re-enabled in the new system.
- Adding the new
stage-b-test-small-2-gpusuite for the TP test.
Overall, this is a solid contribution to improving the CI structure. The changes are correct and align with the stated goals.
alisonshao
force-pushed
the
ci/migrate-lora-tests
branch
3 times, most recently
from
7e32c10 to
3746066
Compare
github-actions
bot
added
the
documentation
alisonshao
force-pushed
the
ci/migrate-lora-tests
branch
from
3746066 to
6a768ea
Compare
alisonshao
requested review from
Fridge003,
Kangyan-Zhou,
ispobock and
merrymercy
as code owners
|
nightly tested at: https://github.com/sgl-project/sglang/actions/runs/20287097914/job/58263191790 |
Part of CI reorganization effort (#13808). Changes: - Move lora_utils.py to python/sglang/test/lora_utils.py - Migrate 6 LoRA test files from test/srt/lora/ to test/registered/lora/ - Add CI registration with appropriate suites: - 5 tests -> stage-b-test-small-1-gpu - 1 test (TP) -> stage-b-test-small-2-gpu - Add stage-b-test-small-2-gpu to per-commit CUDA suites - Remove LoRA entries from old test/srt/run_suite.py - Clean up sys.path hacks, use proper sglang.test imports Not migrated (nightly test): - test_lora_hf_sgl_logprob_diff.py (already registered)
1. Add python/sglang/test/ci/__init__.py to fix ModuleNotFoundError when importing from sglang.test.ci module (missing since PR #13345) 2. Add stage-b-test-2-gpu workflow job to run the stage-b-test-small-2-gpu suite which is used by test_lora_tp.py
- Move test_lora_hf_sgl_logprob_diff.py from test/srt/lora/ to test/registered/lora/ - Remove is_in_ci() skip logic so tests actually run during nightly - Remove entry from __not_in_ci__ section in run_suite.py - Remove outdated comment about nightly test location - Test is registered with suite="nightly-1-gpu" and nightly=True
Moved files (all registered with nightly-1-gpu suite): - test_lora_eviction_policy.py - test_lora_openai_api.py - test_lora_openai_compatible.py - test_lora_qwen3.py - test_lora_radix_cache.py Also fix isort lint issue in test_lora_hf_sgl_logprob_diff.py
Change relative imports to absolute imports: - from lora_utils import ... -> from sglang.test.lora_utils import ... - Remove sys.path.insert hack that was causing ModuleNotFoundError
…t-2-gpu to slash handler - Change partition count from 2 to 3 for stage-b-test-small-1-gpu (was timing out) - Add stage-b-test-2-gpu to slash command handler for rerun-stage support
alisonshao
force-pushed
the
ci/migrate-lora-tests
branch
from
1177e0f to
1404c7e
Compare
…n3_pp * 'main' of https://github.com/sgl-project/sglang: (74 commits) [bug fix][pp] fix inconsistent latency between tp (sgl-project#15379) Fix warp illegal instruction in kimi k2 thinking PCG (sgl-project#15306) Fix gpt-oss yarn with `truncate` argument (sgl-project#14270) Monkey patch deepseek-ocr's `v_head_dim` (sgl-project#15384) [model-gateway] Replace PolicyRegistry RwLock with DashMap for lock-free policy lookups (sgl-project#15361) [PP] Fix dynamic chunking strategy for PP (sgl-project#15372) Fix issue: ENABLE_BELOW_SM90 cannot be enabled on aarch64 CPU (sgl-project#12967) Split test_piecewise_cuda_graph.py to optimize CI resource usage (sgl-project#15290) unified management of environment variables for vlm cuda ipc transport (sgl-project#14501) Mistral Large 3 NVFP4 TRTLLM MoE support (sgl-project#15049) fix: adjust time for test_epd_disaggregation.py (sgl-project#15354) Add doc for qwen3 next (sgl-project#15337) feat: DeepSeek-V3.2 Streaming tool call output (sgl-project#15278) Feature/trtllm mha workspace size configurable sgl-project#15089 (sgl-project#15131) [VLM] Support cos sin cache for Qwen3-VL & GLM-4.1V (sgl-project#15205) [Deepseek V3.2] Support Overlap Spec + NSA (sgl-project#15307) Add request-level timestamp for when prefill finishes (sgl-project#14860) [CI] Migrate LoRA tests to test/registered/lora/ (sgl-project#15176) Reserve more memory for DeepSeekOCR model and adjust server start timeout for DeepGEMM to reduce flakiness (sgl-project#15277) Fix condition check for require_gathered_buffer (sgl-project#15328) ...
2 participants
Summary
Part of CI reorganization effort (#13808).
This PR migrates all LoRA tests to
test/registered/lora/with proper CI registration.Changes
Per-commit tests:
test/srt/lora/totest/registered/lora/stage-b-test-2-gpuworkflow job for 2-GPU testsstage-b-test-small-2-gputoPER_COMMIT_SUITESNightly tests:
test_lora_hf_sgl_logprob_diff.pyfromtest/srt/lora/totest/registered/lora/test/nightly/totest/registered/lora/:is_in_ci()skip logic so nightly tests actually runfrom lora_utils→from sglang.test.lora_utils)Infrastructure fixes:
python/sglang/test/ci/__init__.pyto fixModuleNotFoundError: No module named 'sglang.test.ci'lora_utils.pytopython/sglang/test/lora_utils.pyfor proper importstest/srt/run_suite.pyTest Files Migrated
Per-commit (stage-b-test-small-1-gpu):
Per-commit (stage-b-test-small-2-gpu):
Nightly (nightly-1-gpu):
Test plan
stage-b-test-2-gpujob