Commits · 4b87ffbefb3881a0a33f9c1cb7121429bddad666 · OpenDAS / vllm_cscc

09 Mar, 2026 3 commits

[torch.compile] Rename `compile_ranges_split_points` to `compile_ranges_endpoints` (#36027) · 4b87ffbe

Copilot authored Mar 09, 2026

Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: ProExpertProg <11367180+ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

4b87ffbe

Reapply [Attention] Refactor `check_and_update_config` (#35122) · 77a73458
Matthew Bonanni authored Mar 09, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
77a73458
[Bugfix] Skip out-of-stage layers in get_layers_from_vllm_config for pipeline parallel (#36243) · c4d859c2
Tushar Shetty authored Mar 09, 2026
```
Signed-off-by: Tushar Shetty <tushar.shetty@abbyy.com>
Signed-off-by: Tushar Shetty <54362365+tusharshetty61@users.noreply.github.com>
```
c4d859c2

07 Mar, 2026 2 commits
- [Core] NGram GPU Implementation compatible with Async Scheduler (#29184) · a6be75db
  PatchyTIS authored Mar 08, 2026
  
  a6be75db
- [V0 Deprecation] Remove unused swap_space parameter (#36216) · 00b814ba
  lif authored Mar 07, 2026
```
Signed-off-by: majiayu000 <1835304752@qq.com>
Co-authored-by: mcelrath
```
  00b814ba
06 Mar, 2026 2 commits

[docs][torch.compile] Add fusions.md — kernel/operator fusion reference page (#35538) · ce8546a1

Copilot authored Mar 06, 2026

Signed-off-by: ProExpertProg <luka.govedic@gmail.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: ProExpertProg <11367180+ProExpertProg@users.noreply.github.com>
Co-authored-by: ProExpertProg <luka.govedic@gmail.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

ce8546a1

[Frontend][Core] Add shutdown timeout - allowing in-flight requests to finish (#34730) · 27066d1b
Mark McLoughlin authored Mar 06, 2026
```
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>
```
27066d1b

05 Mar, 2026 1 commit

[Bugfix][CI] fix typos (#34934) · 6a895197

Jiayi Yan authored Mar 06, 2026

Signed-off-by: 1195343015 <1195343015@qq.com>
Signed-off-by: Jiayi Yan <66017932+1195343015@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

6a895197

04 Mar, 2026 1 commit

[Core] Add All-to-All communication backend for DCP (#34883) · 6cb90109

sungsoo ha authored Mar 04, 2026


Signed-off-by: Sungsoo Ha <sungsooh@nvidia.com>
Signed-off-by: sungsoo ha <hasungsoo@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

6cb90109

02 Mar, 2026 1 commit
- [KVConnector] Auto-downgrade to PIECEWISE cudagraph mode for layerwise async ops (#31057) · a13d8c03
  Yashwant Bezawada authored Mar 02, 2026
```
Signed-off-by: Yashwant Bezawada <yashwant_b@me.com>
```
  a13d8c03
01 Mar, 2026 1 commit
- [torch.compile] Undo the fast_moe_cold_start hack in torch>=2.11 (#35475) · e82fbeec
  Richard Zou authored Mar 01, 2026
```
Signed-off-by: Richard Zou <zou3519@gmail.com>
```
  e82fbeec
27 Feb, 2026 2 commits
- [compile] Cleanup: Remove unnecessary +rms_norm forcing for sequence parallelism (#35410) · 66c1751d
  Jason Li authored Feb 27, 2026
```
Signed-off-by: jasonlizhengjian <jasonlizhengjian@gmail.com>
```
  66c1751d
- [Bugfix] disable allreduce_rms_fusion by default when pp size > 1 (#35424) · 487e5c51
  Jiangyun Zhu authored Feb 27, 2026
```
Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
```
  487e5c51
26 Feb, 2026 4 commits

[ROCm] Add extra step in config initialization to populate custom ops before... · 6042e66c

Gregory Shtrasberg authored Feb 26, 2026


[ROCm] Add extra step in config initialization to populate custom ops before compilation config init (#34848)
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>

6042e66c

[torch.compile] Sequence Parallelism threshold compile ranges (#28672) · 9d379410

Jason Li authored Feb 25, 2026


Signed-off-by: jasonlizhengjian <jasonlizhengjian@gmail.com>
Signed-off-by: Jason Li <jasonlizhengjian@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

9d379410

[UX] Add `--performance-mode {balanced,interactivity,throughput}` (#34936) · cbf8f702
Michael Goin authored Feb 25, 2026
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
cbf8f702

[offloader] v2: Hide weight onloading latency via prefetching (#29941) · 6831650c

Ming Yang authored Feb 25, 2026


Signed-off-by: Ming Yang <minos.future@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>

6831650c

25 Feb, 2026 1 commit
- [ROCm]: Enable customop and rope+kvcache fusion for AITER RoPE (#35180) · f38f8c97
  Rohan Potdar authored Feb 24, 2026
```
Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
```
  f38f8c97
24 Feb, 2026 2 commits

[Model][Spec Decode] Nemotron-H MTP and Mamba Speculative Decoding Support (#33726) · f5972a87

Benjamin Chislett authored Feb 24, 2026


Signed-off-by: Shahar Mor <smor@nvidia.com>
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Shahar Mor <smor@nvidia.com>
Co-authored-by: Roi Koren <roik@nvidia.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>

f5972a87

[ROCm] AITER fused RoPE+KVCache (#33443) · 2ff4e511

Rohan Potdar authored Feb 23, 2026


Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
Signed-off-by: charlifu <charlifu@amd.com>
Signed-off-by: Rohan Potdar <66227218+Rohan138@users.noreply.github.com>
Co-authored-by: charlifu <charlifu@amd.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>

2ff4e511

22 Feb, 2026 1 commit
- [Bug] Refactor max_num_batched_tokens to account for drafting (#34898) · 682566b1
  Benjamin Chislett authored Feb 22, 2026
```
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
```
  682566b1
20 Feb, 2026 1 commit
- [CI] Revert PRs 34818 and 33600 (#34979) · aaefc58e
  Lucas Wilkinson authored Feb 20, 2026
  
  aaefc58e
19 Feb, 2026 1 commit

[Bugfix] Fix Basic Models Test (#34818) · 662205d3

Matthew Bonanni authored Feb 19, 2026


Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>

662205d3

18 Feb, 2026 1 commit
- [torch.compile] Turn on silu+fp4 quant fusion by default for O1+ (#34718) · 02e8f26c
  Luka Govedič authored Feb 17, 2026
```
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
```
  02e8f26c
15 Feb, 2026 1 commit

[torch.compile] Disable ar-rms fusion for ds3-fp4 & DP, fix CI test (#34392) · 23d825ab

Luka Govedič authored Feb 15, 2026

Signed-off-by: Luka Govedič <lgovedic@redhat.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

23d825ab

14 Feb, 2026 1 commit
- [Hybrid] Enable mamba prefix cache "align" mode with async scheduling (#33997) · d5fe3f70
  Thomas Parnell authored Feb 14, 2026
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
```
  d5fe3f70
13 Feb, 2026 2 commits
- [Misc] vLLM's --enforce-eager should turn off compile and cudagraphs only (#34523) · 87789c83
  Richard Zou authored Feb 13, 2026
```
Signed-off-by: Richard Zou <zou3519@gmail.com>
```
  87789c83
- [BugFix] Add block_size validation for mamba cache align mode (#34445) · 6f019e6e
  Harry Huang authored Feb 13, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
  6f019e6e
11 Feb, 2026 2 commits

[Docs] Fix typo ("defult") and double spacing (#34348) · 48134a2c

SorenDreano authored Feb 11, 2026

Signed-off-by: SorenDreano <71752785+SorenDreano@users.noreply.github.com>
Co-authored-by: Soren Dreano <soren@numind.ai>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

48134a2c

[torch.compile] Enable AR+rms fusion by default available for `-O2` (#34299) · addac0e6
Luka Govedič authored Feb 11, 2026
```
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
```
addac0e6

08 Feb, 2026 1 commit
- [torch.compile] Add an option to force-enable the MOE cold start optimization (#33735) · 4df841fe
  Richard Zou authored Feb 08, 2026
```
Signed-off-by: Richard Zou <zou3519@gmail.com>
```
  4df841fe
07 Feb, 2026 2 commits

[Kernel] Add KernelConfig flag to enable/disable FlashInfer autotune (#34006) · dd6a6e11

Mohammad Miadh Angkad authored Feb 07, 2026

Signed-off-by: Mohammad Miadh Angkad <176301910+mmangkad@users.noreply.github.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

dd6a6e11

[CI][AMD]Bugfix] Check that model_config is not None in enable_norm_pad_fusion (#34007) · ec28784f
rasmith authored Feb 06, 2026
```
Signed-off-by: Randall Smith <Randall.Smith@amd.com>
```
ec28784f

05 Feb, 2026 3 commits

[Spec Decode] Unified Parallel Drafting (#32887) · af3162d3
Benjamin Chislett authored Feb 05, 2026
```
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
```
af3162d3

[Feat][RL][1/2] Native Weight Syncing API: NCCL (#31943) · c1858b7e

Aaron Hao authored Feb 05, 2026


Signed-off-by: ahao-anyscale <ahao@anyscale.com>
Signed-off-by: Aaron Hao <ahao@anyscale.com>
Co-authored-by: SumanthRH <sumanthrh99@gmail.com>

c1858b7e

[CI][torch.compile] Reduce e2e fusion test time (#33293) · 4d951353

Luka Govedič authored Feb 04, 2026


Signed-off-by: Luka Govedič <lgovedic@redhat.com>
Signed-off-by: ProExpertProg <luka.govedic@gmail.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

4d951353

03 Feb, 2026 1 commit
- Turn `@config` into a `dataclass_transform` (#31541) · 61e632ae
  Harry Mellor authored Feb 03, 2026
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  61e632ae
30 Jan, 2026 3 commits

Disable Cascade Attention for Batch Invariance (#32561) · 8f5d5120

Frank Wang authored Jan 30, 2026

Signed-off-by: frankwang28 <frank.wbb@hotmail.com>
Signed-off-by: Frank Wang <41319051+frankwang28@users.noreply.github.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

8f5d5120

[BugFix] Disable async scheduling for Mamba prefix caching (#33352) · ec51831a
Harry Huang authored Jan 30, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
ec51831a
Fix `tie_word_embeddings` for multimodal models in Transformers v5 (#33359) · 80b918f2
Harry Mellor authored Jan 30, 2026
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
80b918f2