Skip to content

[Deepseek V3.2] Fix Deepseek MTP in V1 mode#15429

Merged
Fridge003 merged 1 commit intosgl-project:mainfrom
bzhng-development:brayden/fix-deepseek-nsa-mtp
Dec 19, 2025
Merged

[Deepseek V3.2] Fix Deepseek MTP in V1 mode#15429
Fridge003 merged 1 commit intosgl-project:mainfrom
bzhng-development:brayden/fix-deepseek-nsa-mtp

Conversation

@b8zhong
Copy link
Collaborator

@b8zhong b8zhong commented Dec 19, 2025

Motivation

Screenshot 2025-12-18 at 4 51 15 PM

Modifications

Restore the earlier behaviour of Deepseek V3 + MTP. (Accidental change)

Fix #15428

@gemini-code-assist
Copy link
Contributor

Warning

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

@b8zhong b8zhong changed the title [Deepseek V3.2] Fix [Deepseek V3.2] Fix Deepseek MTP in V1 mode Dec 19, 2025
@b8zhong
Copy link
Collaborator Author

b8zhong commented Dec 19, 2025

/tag-and-rerun-ci

# were accepted.
page_table = torch.repeat_interleave(
page_table, repeats=extend_seq_lens_cpu, dim=0
page_table, repeats=forward_batch.extend_seq_lens, dim=0
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

may we add a unit test for this, so that we will not break next time

Copy link
Collaborator

@Fridge003 Fridge003 Dec 19, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We have a per-commit unit test for MTP and the prior PR passed
https://github.com/sgl-project/sglang/actions/runs/20293560819/job/58356702062?pr=15307
Seems this bug only happens during deepgemm precompile

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

good to know

@Fridge003 Fridge003 merged commit e88e75a into sgl-project:main Dec 19, 2025
139 of 151 checks passed
@b8zhong b8zhong deleted the brayden/fix-deepseek-nsa-mtp branch December 19, 2025 03:58
xiaobaicxy added a commit to xiaobaicxy/sglang that referenced this pull request Dec 19, 2025
* 'main' of https://github.com/sgl-project/sglang: (136 commits)
  fix: unreachable error check in retraction (sgl-project#15433)
  [sgl-kernel] chore: update deepgemm version (sgl-project#13402)
  [diffusion] multi-platform: support diffusion on amd and fix encoder loading on MI325 (sgl-project#13760)
  [amd] Add deterministic all-reduce kernel for AMD (ROCm) (sgl-project#15340)
  [diffusion] refactor: refactor _build_req_from_sampling to use shallow_asdict (sgl-project#13782)
  Add customized sampler registration (sgl-project#15423)
  Update readme (sgl-project#15425)
  Fix Mindspore model import warning (sgl-project#15287)
  [Feature] Xiaomi `MiMo-V2-Flash` day0 support (sgl-project#15207)
  [diffusion] profiling: add bench_serving.py and VBench (sgl-project#15410)
  [DLLM] Fix dLLM regression (sgl-project#15371)
  [Deepseek V3.2] Fix Deepseek MTP in V1 mode (sgl-project#15429)
  chore: update CI_PERMISSIONS (sgl-project#15431)
  [DLLM] Add CI for diffusion LLMs (sgl-project#14723)
  Support using different attention backend for draft decoding. (sgl-project#14843)
  feat(dsv32): better error handling for DeepSeek-v3.2 encoder (sgl-project#14353)
  tiny fix lint on main (sgl-project#15424)
  multimodal: precompute hash for MultimodalDataItem (sgl-project#14354)
  [AMD] Clear pre-built AITER kernels and warmup to prevent segfaults and test timeouts (sgl-project#15318)
  [Performance] optimize NSA backend metadata computation for multi-step speculative decoding (sgl-project#14781)
  ...
Prozac614 pushed a commit to Prozac614/sglang that referenced this pull request Dec 23, 2025
jiaming1130 pushed a commit to zhuyijie88/sglang that referenced this pull request Dec 25, 2025
YChange01 pushed a commit to YChange01/sglang that referenced this pull request Jan 13, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Bug] TypeError in NSA backend: torch.repeat_interleave called with repeats=list during DeepSeek-V3.2 DEEPGEMM warm up (nightly docker)

3 participants