Skip to content

[Misc]feat: adapt to vLLM main (0d4d334e...46a95815) - #245

Draft
Meihan-chen wants to merge 36 commits into
nv-action:mainfrom
Meihan-chen:main2main_auto_2026-05-18_16-58
Draft

[Misc]feat: adapt to vLLM main (0d4d334e...46a95815)#245
Meihan-chen wants to merge 36 commits into
nv-action:mainfrom
Meihan-chen:main2main_auto_2026-05-18_16-58

Conversation

@Meihan-chen

@Meihan-chen Meihan-chen commented May 18, 2026 •
edited by github-actions Bot
Loading

Copy link
Copy Markdown
Collaborator

Main2Main Summary

Status: completed
Upstream range: 0d4d334eaa583b9c09aa4eb7538c22db99fd84b3..46a95815d344e43ebd6b4757632c55d03e90983a
Reached upstream commit: 46a95815d344e43ebd6b4757632c55d03e90983a
Steps: 8/8
CI suite: e2e-singlecard-light

Result

The target upstream commit 46a95815d344e43ebd6b4757632c55d03e90983a was fully reached. All 8 planned steps were verified through CI, committed with signed commits, and the vLLM commit reference was updated at each step. One code adaptation was required (step-5: token_arange_np dtype), while all other steps were commit reference updates only.

Completed Steps

Step Upstream range vllm-ascend commit CI result Summary
step-1 0d4d334..27b85d2 381281b passed InternS2Preview model, CPU/RISC-V attention, thinking_budget fix (commit reference only)
step-2 27b85d2..2676ab1 357abba passed Deprecated shim removal for get_tokenizer/resolve_hf_chat_template (commit reference only)
step-3 2676ab1..4b364f8 d8fbcbf passed LoRA maybe_get_oot_by_class, KV cache SWA optimization (commit reference only)
step-4 4b364f8..75fd68c 3777e1b passed Pooling offline API refactor (commit reference only)
step-5 75fd68c..0fe7550 a094d1c passed Added dtype=np.int32 to token_arange_np in eagle_proposer.py and tests
step-6 0fe7550..e0a45f1 c042924 passed RL IPC weight sync optimizations (commit reference only)
step-7 e0a45f1..fb5bd03 4d2d4ca passed kv_connector ordering fix, IR op priority to worker init (commit reference only)
step-8 fb5bd03..46a9581 c3e27b8 passed ROCm MLA FP8 ASM prefill (commit reference only)

Changes Made

  • Updated vLLM commit reference from 0d4d334eaa583b9c09aa4eb7538c22db99fd84b3 to 46a95815d344e43ebd6b4757632c55d03e90983a in:
    • docs/source/conf.py
    • .github/workflows/dockerfiles/Dockerfile.lint
    • .github/workflows/pr_test_full.yaml
    • .github/workflows/pr_test_light.yaml
  • Adapted vllm_ascend/spec_decode/eagle_proposer.py: Added dtype=np.int32 to token_arange_np for consistency with upstream SpecDecodeBaseProposer
  • Updated tests/ut/spec_decode/test_eagle_proposer.py: Updated test token_arange_np assignments to match dtype

CI Verification

  • Passed: All 8 steps (step-1 through step-8) with e2e-singlecard-light suite
  • Treated as env flakes: none
  • Last successful step: step-8 (target commit reached)

Follow-up

None required. All upstream changes absorbed successfully. The upgrade is complete and verified through CI.

Meihan-chen and others added 30 commits May 18, 2026 21:48
@Meihan-chen
fix yaml
e4ba56c 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
main2main_auto
6f54c70 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
main2main_auto_test
6bac7b4 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
refactor bisect
9379713 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
refactor issue review
763cf22 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
docs: add workflow-native main2main orchestration spec
6e4fe80
@Meihan-chen
refactor: migrate main2main to workflow-native orchestration
b9ead70
@Meihan-chen
fix: harden workflow-native main2main guards
b3b06bc
@Meihan-chen
fix
1898729 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
fix 2
12502ee 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
fix 3
2c69a6b 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
fix 3
fa2e535 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
feat: add simplified main2main workflows
a2c2703
@zhangxinyuehfad @Meihan-chen
fix doc
60a0b0a 
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
@Meihan-chen
main2main_auto_fix
9795c16 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@zhangxinyuehfad @Meihan-chen
fix doc
fd76eb8 
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
@Meihan-chen
main2main_auto_fix
bedae6c 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
main2main_auto_fix
ec0ed6e 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
main2main_mock
693f068 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
main2main_refactor
08a2b90 
Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
@Meihan-chen
fix skill
2f4ce70 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
fix main2main
aba803a 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main auto
7ee2eb8 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main skill
99ae085 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main auto
50338b4 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main auto claude action
c4fe043 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main auto claude cli
de24493 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
@Meihan-chen
main2main auto
2b167ec 
Signed-off-by: Meihan-chen <239345480+Meihan-chen@users.noreply.github.com>
main2main-bot
main2main: absorb vLLM commits 0d4d334e..27b85d2084
381281b 
Upstream changes absorbed:
- InternS2Preview model support (new model registry entry)
- CPU/RISC-V attention backend (no Ascend impact)
- Bug fix for thinking_budget_tracks_reqs condition in gpu_input_batch
- Speculative config handling for intern_s2_preview

vllm-ascend impact: No code changes required (no-op adapt).
Commit reference updated in conf.py and workflow files.

Signed-off-by: main2main-bot <main2main-bot@users.noreply.github.com>
main2main-bot
main2main: absorb vLLM commit 2676ab1e (deprecation removal)
357abba 
Upstream changes absorbed:
- Removed deprecated shim for vllm.transformers_utils.tokenizer.get_tokenizer
- Removed deprecated shim for vllm.entrypoints.chat_utils.resolve_hf_chat_template

vllm-ascend impact: No code changes required (no-op adapt).
vllm-ascend doesn't import from the deprecated locations.
Commit reference updated in conf.py and workflow files.

Signed-off-by: main2main-bot <main2main-bot@users.noreply.github.com>
@zhangxinyuehfad
zhangxinyuehfad force-pushed the main branch 5 times, most recently from de24f70 to 60cd7c9 Compare June 18, 2026 04:04
@zhangxinyuehfad
zhangxinyuehfad force-pushed the main branch 11 times, most recently from 05b98c1 to 38085d5 Compare June 26, 2026 09:09
@zhangxinyuehfad
zhangxinyuehfad force-pushed the main branch 6 times, most recently from 1cd9ff4 to b7570c6 Compare July 7, 2026 03:28
@zhangxinyuehfad
zhangxinyuehfad force-pushed the main branch 4 times, most recently from 0c96baa to e7d055b Compare July 13, 2026 09:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@wangxiyuan wangxiyuan Awaiting requested review from wangxiyuan wangxiyuan will be requested when the pull request is marked ready for review wangxiyuan is a code owner
@Yikun Yikun Awaiting requested review from Yikun Yikun will be requested when the pull request is marked ready for review Yikun is a code owner

Assignees

No one assigned

Labels

ci/build documentation Improvements or additions to documentation merge-conflicts module:tests

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@Meihan-chen @zhangxinyuehfad