Commits · 8aa1485fcff7be3e42300c0615ee0f3f3cbce9a8 · OpenDAS / vllm_cscc · GitLab

28 Jul, 2025 38 commits
- [Perf] Disable chunked local attention by default with llama4 (#21761) · 8aa1485f
  Lucas Wilkinson authored Jul 28, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  8aa1485f
- [Feat]: Add support for Dynamic Quant 4 bit CPU kleidiai kernels (#17112) · 89ac266b
  Nikhil Gupta authored Jul 28, 2025
```
Signed-off-by: Nikhil Gupta <nikhil.gupta2@arm.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
```
  89ac266b
- [Bugfix] DeepGEMM is not enabled on B200 due to `_lazy_init()` (#21472) · c6f36cfa
  Clayton Coleman authored Jul 28, 2025
```
Signed-off-by: Clayton Coleman <smarterclayton@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
```
  c6f36cfa
- Revert "[V1] Exception Handling when Loading KV Cache from Remote Store" (#21778) · b18b417f
  Kuntai Du authored Jul 28, 2025
```
Signed-off-by: KuntaiDu <kuntai@uchicago.edu>
```
  b18b417f
- [Bug] Enforce contiguous input for `dynamic_scaled_fp8_quant` and... · e0e58f97
  Wentao Ye authored Jul 28, 2025
```
[Bug] Enforce contiguous input for `dynamic_scaled_fp8_quant` and `static_scaled_fp8_quant` (#21773)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  e0e58f97
- [AMD][BugFix] Fix omission of wvSplitK kernel for small batch sizes (1-4) due... · b361f14e
  rasmith authored Jul 28, 2025
```
[AMD][BugFix] Fix omission  of wvSplitK kernel for small batch sizes (1-4) due to torch.compile (#21350)
Signed-off-by: Randall Smith <Randall.Smith@amd.com>
```
  b361f14e
- Use `metavar` to list the choices for a CLI arg when custom values are also accepted (#21760) · 94b71ae1
  Harry Mellor authored Jul 28, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  94b71ae1
- [P/D] Log warnings related to prefill KV expiry (#21753) · 7d44c691
  Nick Hill authored Jul 28, 2025
```
Signed-off-by: Nick Hill <nhill@redhat.com>
```
  7d44c691
- [Bugfix] Fix granite speech shape validation (#21762) · e17a4d3b
  Cyrus Leung authored Jul 29, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  e17a4d3b
- [XPU] IPEX-optimized Punica Wrapper on XPU (#21703) · ec261b02
  Chaojun Zhang authored Jul 29, 2025
```
Signed-off-by: chzhang <chaojun.zhang@intel.com>
Co-authored-by: Jee Jee Li <pandaleefree@gmail.com>
```
  ec261b02
- [Bugfix] Mistral crashes on tool with no description (#21167) · 25708d31
  Michard Hugo authored Jul 28, 2025
```
Signed-off-by: HugoMichard <hugo@harfanglab.fr>
```
  25708d31
- [Misc] Reduce logs for model resolution (#21765) · 0e18a5d0
  Cyrus Leung authored Jul 28, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  0e18a5d0
- [Logs] Change flashinfer sampler logs to once (#21759) · 34a20c49
  Michael Goin authored Jul 28, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  34a20c49
- [Bugfix]check health for engine core process exiting unexpectedly (#21728) · bccc43c0
  wuhang authored Jul 28, 2025
```
Signed-off-by: wuhang <wuhang6@huawei.com>
```
  bccc43c0
- [Bugfix] Improve JSON extraction in LlamaToolParser (#19024) · 9ace2eaf
  Keyang Ru authored Jul 28, 2025
```
Signed-off-by: keru <keyang.ru@oracle.com>
Co-authored-by: keru <keyang.ru@oracle.com>
```
  9ace2eaf
- [`Ernie 4.5`] Name Change for Base 0.3B Model (#21735) · 656c24f1
  Anton Vlasjuk authored Jul 28, 2025
```
Signed-off-by: vasqu <antonprogamer@gmail.com>
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  656c24f1
- [Bugfix] Fix glm4.1v video_grid_thw tensor shape scheme (#21744) · 0ae970ed
  Isotr0py authored Jul 28, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  0ae970ed
- [Bugfix] Fix Ernie4_5_MoeForCausalLM shared experts (#21717) · 1b769dcc
  Jee Jee Li authored Jul 28, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  1b769dcc
- [feature] add log non default args in LLM (#21680) · 2cc57119
  rongfu.leng authored Jul 28, 2025
```
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
```
  2cc57119
- [Model] Prioritize Transformers fallback over suffix matching (#21719) · a4ed7315
  Cyrus Leung authored Jul 28, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  a4ed7315
- Migrate KeyeImageInputs and KeyeVideoInputs to TensorSchema (#21686) · d128d0d5
  Benji Beck authored Jul 28, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
```
  d128d0d5
- [v1][mamba] Added mamba_type into MambaSpec (#21715) · a6c05028
  Asaf Joseph Gardin authored Jul 28, 2025
```
Signed-off-by: asafg <asafg@ai21.com>
Co-authored-by: asafg <asafg@ai21.com>
```
  a6c05028
- [BugFix] Fix ChunkedLocalAttention when the hybrid kv-cache is disabled (#21707) · 139a7f07
  Lucas Wilkinson authored Jul 28, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  139a7f07
- [Bugfix] Fix shape checking for Fuyu (#21709) · 139a97ec
  Cyrus Leung authored Jul 28, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  139a97ec
- [bugfix] fix profile impact benchmark results (#21507) · 18cc33dd
  rongfu.leng authored Jul 28, 2025
```
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
```
  18cc33dd
- [Bugfix] [issue-21565] Fix the incompatibility issue with stream and named... · 7656cf4c
  Hongsheng Liu authored Jul 28, 2025
```
[Bugfix] [issue-21565] Fix the incompatibility issue with stream and named function calling when Thinking is disabled (#21573)
Signed-off-by: wangzi <3220100013@zju.edu.cn>
Co-authored-by: wangzi <3220100013@zju.edu.cn>
```
  7656cf4c
- Migrate Idefics3ImagePixelInputs and Idefics3ImageEmbeddingInputs to … (#21683) · 3ea57a56
  Benji Beck authored Jul 27, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
```
  3ea57a56
- Migrate GraniteSpeechAudioInputs to TensorSchema (#21682) · 75856bc2
  Benji Beck authored Jul 27, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
```
  75856bc2
- Migrate GLMVImagePixelInputs to TensorSchema (#21679) · 304dcdf5
  Benji Beck authored Jul 27, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
```
  304dcdf5
- Migrate Glm4vImageInputs, Glm4vVideoInputs to TensorSchema (#21678) · 88e46c7c
  Benji Beck authored Jul 27, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
```
  88e46c7c
- Migrate Gemma3ImagePixelInputs to TensorSchema (#21676) · d8937de4
  Benji Beck authored Jul 27, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
```
  d8937de4
- [FEAT] [ROCm] [AITER]: Add AITER HIP block quant kernel (#21242) · e626d286
  TJian authored Jul 27, 2025
  
  e626d286
- [Model] Support TP/PP/mamba2 kernel for PLaMo2 (#19674) · c7ffe93d
  Shinichi Hemmi authored Jul 28, 2025
```
Signed-off-by: Shinichi Hemmi <shemmi@preferred.jp>
Signed-off-by: Shinichi Hemmi <50256998+Alnusjaponica@users.noreply.github.com>
Co-authored-by: Calvin Metzger <metzger@preferred.jp>
Co-authored-by: Sixue Wang <cecilwang@preferred.jp>
```
  c7ffe93d
- [V1] Exception Handling when Loading KV Cache from Remote Store (#21534) · 15a72ac4
  Adeline authored Jul 28, 2025
```
Signed-off-by: liuyumoye <adeline_ly2023@outlook.com>
Co-authored-by: liuyumoye <adeline_ly2023@outlook.com>
```
  15a72ac4
- [Misc] Add fused_moe configs for Qwen3-Coder-480B-A35B-Instruct-FP8 (#21700) · 04ff4be3
  Jee Jee Li authored Jul 28, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  04ff4be3
- Fix GLM tool parser (#21668) · 93269bb4
  Yuxuan Zhang authored Jul 28, 2025
```
Co-authored-by: Chenhui Zhang <zhang.chenhui@outlook.com>
```
  93269bb4
- Fix typo for limit-mm-per-prompt in docs (#21697) · 82acf218
  Joachim Studnia authored Jul 27, 2025
```
Signed-off-by: Joachim Studnia <joachim@mistral.ai>
```
  82acf218
- [Deprecation][2/N] Replace `--task` with `--runner` and `--convert` (#21470) · 86ae693f
  Cyrus Leung authored Jul 28, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  86ae693f
27 Jul, 2025 2 commits
- [Attention] Make CutlassMLA the default backend for SM100 (blackwell) (#21626) · 8f605ee3
  Alexander Matveev authored Jul 27, 2025
```
Signed-off-by: Alexander Matveev <amatveev@redhat.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
```
  8f605ee3
- [Misc] Refactor vllm config str (#21666) · a9b2a1d7
  Ning Xie authored Jul 28, 2025
  
  a9b2a1d7