[Spec] Deprecate Spec V1#25464
Merged
Merged
Conversation
hnyls2002
requested review from
ByronHsu,
Qiaolin-Yu,
ShangmingCai,
Ying1123,
iforgetmyname,
kpham-sgl,
merrymercy,
ping1jing2 and
xiezhq-hermann
as code owners
May 16, 2026 05:25
Contributor
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
Collaborator
Author
|
/tag-and-rerun-ci |
hnyls2002
changed the title
hnyls2002
force-pushed
the
lsyin/draft-executor-abc
branch
from
May 16, 2026 21:17
be855fd to
f638592
Compare
|
hopefully it will be merged soon |
hnyls2002
force-pushed
the
lsyin/draft-executor-abc
branch
from
May 31, 2026 08:51
f638592 to
a847b66
Compare
hnyls2002
requested review from
hanming-lu,
hzh0425,
ispobock and
yizhang2077
as code owners
May 31, 2026 09:01
hnyls2002
force-pushed
the
lsyin/draft-executor-abc
branch
2 times, most recently
from
June 8, 2026 09:31
c07f055 to
546285f
Compare
hnyls2002
force-pushed
the
lsyin/draft-executor-abc
branch
from
June 8, 2026 10:31
d15c28f to
1e9b41e
Compare
hnyls2002
changed the title
Oasis-Git
added a commit
to Oasis-Git/sglang
that referenced
this pull request
Jun 8, 2026
7 commits from main. Notable: - sgl-project#27528 Fix GPT-OSS MXFP4 hidden size reshape on SM10X — upstream landed the same fix I had locally for the prepad bug in forward_normal; dropped my local commit f3c169b85 before merging so the upstream version comes in cleanly. - sgl-project#25464 Deprecate Spec V1 — removed python/sglang/srt/speculative/{eagle_worker,multi_layer_eagle_worker, standalone_worker}.py. We had local edits on eagle_worker.py for the cuda_graph_bs -> cuda_graph_bs_decode rename; accepted the deletion (V2 was already the runtime path; no remaining call sites for V1 in srt/). - Other content (dsv4 ROCm fix, EAGLE SWA pool, TRTLLM target verify, NPU CI, LMSYS blog) auto-merged without conflicts. Stale-import surface re-checked post-merge: no leftover imports of piecewise_context_manager, the old piecewise_cuda_graph_runner, the old cuda_graph_runner module, or the deleted V1 spec workers. No stale server_args.cuda_graph_bs / disable_piecewise_cuda_graph accesses in python/sglang/srt/. Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>
jeynmann
pushed a commit
to jeynmann/sglang
that referenced
this pull request
Jun 9, 2026
This was referenced Jun 10, 2026
This was referenced Jun 11, 2026
JustinTong0323
added a commit
to JustinTong0323/sglang
that referenced
this pull request
Jun 14, 2026
The Spec V1 deprecation (sgl-project#25464) deleted eagle_worker.py, which carried the only setter for forward_batch.reuse_mtp_topk_indices. The receiver plumbing (deepseek_nextn, ForwardBatch fields, TBO filter) from sgl-project#27114 survived but went dead. Re-add the setup/reset block to the V2 EagleDraftWorker.draft_forward so DSA/NSA topk indices are shared across MTP draft steps again (topk==1 only).
jmamou
added a commit
to jmamou/sglang
that referenced
this pull request
Jun 17, 2026
standalone_worker.py was deleted upstream (Spec V1 deprecated in sgl-project#25464). Port _validate_vocab_compatibility to StandaloneWorkerV2 in standalone_worker_v2.py and update the test accordingly.
timmy-feng
added a commit
to modal-projects/sglang
that referenced
this pull request
Jun 26, 2026
5 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
eagle_worker.py,multi_layer_eagle_worker.py,standalone_worker.py(~2.3k lines deleted)--disable-overlap-schedule) path drives the same V2 worker synchronouslySpeculativeAlgorithm.create_workerand the implicitis_spec_v2 == overlapcoupling —is_spec_v2now means "uses the V2 worker/schema", independent of scheduling modeBackground — why V1 can be removed now
V1 was retained only because the V2 path lacked feature parity. The remaining gaps have since landed on main:
topk > 1) on V2: page_size==1 (Support spec v2 tree drafting (eagle topk>1) with page_size==1 #26866, relanded as Reland spec v2 tree drafting (eagle topk>1) with page_size==1 (#26866) #26997), page_size>1 (Spec v2 tree drafting (topk>1) with page_size>1 #26972), and mamba / hybrid-linear models (Supporttopk > 1tree drafting for mamba/hybrid-linear models on spec v2 #27463)topk > 1(Fix NaN in triton EAGLE spec-v2 draft-extend CUDA graph at topk>1 (wrong qo_indptr stride) #27545)With parity in place and the scheduler able to drive the V2 worker on the non-overlap path, the V1 implementations are redundant.
Non-overlap (synchronous) V2 path
Scheduler.run_batchgains a synchronous V2 branch: resolve inputs, runforward_batch_generationinside a forward-isolation context, then re-applyspec_info/seq_lensand copy results to CPU_overlap_forward_isolationgeneralized to_forward_isolation(batch, *, overlap); the non-overlap path skips the 2-iter tensor pinning (single stream, nobatch_record_buf)Test
test_spec_eagle.py,test_spec_eagle_triton.py,test_spec_eagle_topk.py,test_spec_standalone.py,test_spec_eagle_parity.pytest_streaming_session_extra.pyCI States
Latest PR Test (Base): ❌ Run #27131714984
Latest PR Test (Extra): ❌ Run #27131714815