Commits · 7d8975de84dadb87df23eae4a663a3c6620433dd · OpenDAS / vllm_cscc

"docs/deployment/frameworks/cerebrium.md" did not exist on "b25cfab9a03b5c460fb92340b310d2a5c2dbc5da"

16 Oct, 2025 5 commits

Deepseek-v3 Batch Invariant on 8xH100 (#26609) · 7d8975de

Bram Wasti authored Oct 15, 2025


Signed-off-by: Bram Wasti <bwasti@meta.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

7d8975de

[PERF] Qwen3-next MTP speedup (change bool mask indexing to index_select /... · 785d8b64

Vadim Gimpelson authored Oct 16, 2025


[PERF] Qwen3-next MTP speedup (change bool mask indexing to index_select / index_copy to reduce d2h) (#26437)
Signed-off-by: Vadim Gimpelson <vadim.gimpelson@gmail.com>

785d8b64

[Chore] Rename `utils` submodules (#26920) · f6cdc9a0
Cyrus Leung authored Oct 16, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
f6cdc9a0

[ROCm][FEAT] Fuse DeepSeek shared experts into AITER fused_moe ops (#24097) · 13170343

kliuae authored Oct 16, 2025


Signed-off-by: chenjun <junchen2@amd.com>
Signed-off-by: kliuae <kuanfu.liu@embeddedllm.com>
Co-authored-by: valarLip <103567126+valarLip@users.noreply.github.com>
Co-authored-by: TJian <tunjian.tan@embeddedllm.com>

13170343

[Qwen3-Next] Add tuned MoE config for Qwen3-Next FP8 on H100 tp2 (#26887) · f96bc364
felixzhu555 authored Oct 15, 2025
```
Signed-off-by: Felix Zhu <felixzhu555@gmail.com>
```
f96bc364

15 Oct, 2025 7 commits

support flashinfer_fp4 moe for 5090 gpu (#26669) · 0b99f5d3

XiaobingZhang authored Oct 16, 2025


Signed-off-by: XiaobingSuper <xiaobingzhangupc@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>

0b99f5d3

[NVIDIA] Add support for cudnn fp4 gemm via flashinfer (#26107) · de92d916

Kaixi Hou authored Oct 15, 2025


Signed-off-by: kaixih <kaixih@nvidia.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

de92d916

[ModelOpt] Remove NVFP4 MoE K%16==0 constraint (#26891) · d7963752
XiaobingZhang authored Oct 16, 2025
```
Signed-off-by: XiaobingSuper <xiaobingzhangupc@gmail.com>
```
d7963752
[Chore] Separate out `vllm.utils.func` (#26904) · 136a17fe
Cyrus Leung authored Oct 15, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
136a17fe
[Misc] Remove `isort` and `yapf` ignores (#26888) · f93e3480
Cyrus Leung authored Oct 15, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
f93e3480
[Model][2/N] Improve all pooling task | Support multi-vector retrieval (#25370) · f54f8512
wang.yuqi authored Oct 15, 2025
```
Signed-off-by: wang.yuqi <noooop@126.com>
```
f54f8512

[Frontend][torch.compile] CompilationConfig Overhaul (#20283): name change ... · 96b9aa5a

Morrison Turnansky authored Oct 14, 2025


[Frontend][torch.compile] CompilationConfig Overhaul (#20283): name change  compilation level to compilation mode, deprecation compilation level (#26355)
Signed-off-by: morrison-turnansky <mturnans@redhat.com>
Signed-off-by: Morrison Turnansky <mturnans@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

96b9aa5a

14 Oct, 2025 7 commits
- [CI Failure] Fix torchao dep failure for Quantization Test (#26824) · 7e0ef408
  Michael Goin authored Oct 14, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  7e0ef408
- Added MoE configs for llama 4, H200 device with tp=4/8 tuning (#26837) · 0e658189
  Dhruvil Bhatt authored Oct 14, 2025
```
Signed-off-by: Dhruvil Bhatt <bhattdbh@amazon.com>
```
  0e658189
- llama4_vision_rope: add HIP override to accept (q, k) and avoid (positions, q, k) mismatch (#26790) · 87efc681
  Huamin Li authored Oct 14, 2025
```
Signed-off-by: Huamin Li <3ericli@gmail.com>
```
  87efc681
- [Kernel][MoE] Add MoE tunings for GLM 4.6-FP8 and GLM 4.5 Air on NVidia B200 (#26818) · aba48f7d
  Ze'ev Klapow authored Oct 14, 2025
  
  aba48f7d
- [Feature][Quantization] auto_round format add support for regex (#24024) · 29350922
  Heng Guo authored Oct 14, 2025
```
Signed-off-by: n1ck-guo <heng.guo@intel.com>
Signed-off-by: Heng Guo <heng.guo@intel.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  29350922
- [torch.compile] Unwrap fused_marlin_moe custom op (#26739) · 8ae16928
  Varun Sundar Rabindranath authored Oct 13, 2025
```
Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
```
  8ae16928
- [UX] Replace VLLM_ALL2ALL_BACKEND with --all2all-backend (#26732) · 3e051bda
  Michael Goin authored Oct 13, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  3e051bda
13 Oct, 2025 5 commits
- [Bug] Fix Assertion error DeepEP/csrc/kernels/intranode.cu:928: 'false and... · 7200a21c
  Wentao Ye authored Oct 13, 2025
```
[Bug] Fix Assertion error DeepEP/csrc/kernels/intranode.cu:928: 'false and Unsupported type' (#26532)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  7200a21c
- [Quantization] [Performance] Enable Marlin GEMM kernels for the... · 89342ce4
  Alex Kogan authored Oct 13, 2025
```
[Quantization] [Performance] Enable Marlin GEMM kernels for the calibration-free RTN-based quantization (#26051)
Signed-off-by: Alex Kogan <alex.kogan@oracle.com>
Signed-off-by: Alex Kogan <82225080+sakogan@users.noreply.github.com>
```
  89342ce4
- [CI/Build] upgrade compressed-tensors to 0.12.2 to address LGPLv3 (#26501) · a1b2d658
  Sangyeon Cho authored Oct 14, 2025
```
Signed-off-by: Sangyeon Cho <josang1204@gmail.com>
```
  a1b2d658
- [unrevert] Add batch invariant kernel override for FlashInfer backend [2/n] (#26373) · 32637990
  Bram Wasti authored Oct 13, 2025
```
Signed-off-by: Bram Wasti <bwasti@meta.com>
Signed-off-by: Bram Wasti <bwasti@fb.com>
```
  32637990
- [Misc] cache result of disable_inplace (#26666) · 60e419c1
  bnellnm authored Oct 12, 2025
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  60e419c1
12 Oct, 2025 3 commits

Update `Optional[x]` -> `x | None` and `Union[x, y]` to `x | y` (#26633) · 8fcaaf6a
Harry Mellor authored Oct 12, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
8fcaaf6a
[Bugfix][CI/Build] Fix failing Mteb CI (#26638) · 045b396d
Isotr0py authored Oct 12, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
045b396d

[PERF] [Qwen3-next] Speed up gated RMSNorm (#26207) · 82e64c7a

Vadim Gimpelson authored Oct 12, 2025

Signed-off-by: Vadim Gimpelson <vadim.gimpelson@gmail.com>
Signed-off-by: Vadim Gimpelson <156319763+vadiklyutiy@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

82e64c7a

11 Oct, 2025 2 commits

[Refactor]: Use M-RoPE interface directly while defining model class instead... · 727144be

dsinghvi authored Oct 11, 2025


[Refactor]: Use M-RoPE interface directly while defining model class instead of maintaining model specific M-RoPE implementation in mrope.py (#24172)
Signed-off-by: Divyansh Singhvi <divyanshsinghvi@gmail.com>
Signed-off-by: dsinghvi <divyanshsinghvi@gmail.com>
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: wwl2755 <wangwenlong2755@gmail.com>

727144be

[MM] Move Qwen3Omni MRoPE impl to model file (#26608) · ddaff293

Roger Wang authored Oct 10, 2025


Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

ddaff293

10 Oct, 2025 5 commits

Fix some typing issues found by `mypy==1.18.2` (#26596) · 7c12763b
Harry Mellor authored Oct 10, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
7c12763b

Add Qwen3-Omni moe thinker (#25550) · 19a9b169

Xiong Wang authored Oct 11, 2025


Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Roger Wang <hey@rogerw.io>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Xiong Wang <feizi.wx@alibaba-inc.com>
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

19a9b169

[Transform] [Quantization] Add QuTLASS support to vLLM (#24440) · 96ad65b7

Roberto L. Castro authored Oct 10, 2025


Signed-off-by: LopezCastroRoberto <roberto.lopez.castro@udc.es>
Signed-off-by: Roberto L. Castro <38211239+LopezCastroRoberto@users.noreply.github.com>
Signed-off-by: Andrei Panferov <andrei@panferov.org>
Co-authored-by: Andrei Panferov <andrei@panferov.org>
Co-authored-by: Michael Goin <mgoin64@gmail.com>

96ad65b7

Silu v2 (#25074) · 7b03584d

Elvir Crnčević authored Oct 10, 2025


Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: elvircrn <elvircrn@gmail.com>
Signed-off-by: Elvir Crnčević <elvircrn@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com>

7b03584d

[Bugfix] Convert untraceable GroupShape to list for AMD impl (#26535) · 213b6445
Lucas Kabela authored Oct 10, 2025
```
Signed-off-by: Lucas Kabela <lucaskabela@meta.com>
```
213b6445

09 Oct, 2025 6 commits
- [Bug] Fix modular_kernel: ZeroDivisionError: integer division or modulo by zero (#26528) · 1ee35382
  Wentao Ye authored Oct 09, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  1ee35382
- [Bugfix] Disable moe inplace for torch >= 2.9 (#26497) · a462331e
  bnellnm authored Oct 09, 2025
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  a462331e
- [Model] Apply shared experts overlap optimization to all models with shared experts (#26145) · 47e66c24
  bnellnm authored Oct 09, 2025
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  47e66c24
- [V0 deprecation] Remove `QKVCrossParallelLinear` implementation (#26475) · d1ddf340
  Isotr0py authored Oct 09, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  d1ddf340
- [torchao] Add support for ModuleFqnToConfig using regex (#26001) · a83ff278
  Jerry Zhang authored Oct 09, 2025
```
Signed-off-by: Jerry Zhang <jerryzh168@gmail.com>
```
  a83ff278
- Separate MLAAttention class from Attention (#25103) · e614ab78
  Naveenraj Kamalakannan authored Oct 08, 2025
```
Signed-off-by: Naveenraj Kamalakannan <therealnaveenkamal@gmail.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
```
  e614ab78