Commits · 19b927e52df8400084df1c8116af7d6f0a5f5d15 · OpenDAS / vllm_cscc · GitLab

13 Aug, 2025 9 commits
- [Core] Use individual MM items in P0/P1 cache and model runner (#22570) · 19b927e5
  Cyrus Leung authored Aug 13, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  19b927e5
- [Nixl][CI] Fix tests (#22806) · 6b794c75
  Nicolò Lucchesi authored Aug 13, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  6b794c75
- [Frontend] Add chunked processing to handle long inputs in embedding models (#22280) · 653124bd
  Kdump authored Aug 13, 2025
```
Signed-off-by: x22x22 <wadeking@qq.com>
Signed-off-by: Kdump <rootshellexp@gmail.com>
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
Co-authored-by: Maximilien de Bayser <maxdebayser@gmail.com>
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  653124bd
- [V1] Add tree drafting tests for eagle spec decoding (#22705) · d94e3026
  Giancarlo Delfin authored Aug 13, 2025
```
Signed-off-by: Giancarlo Delfin <gdelfin@meta.com>
```
  d94e3026
- [Bugfix] Fix Nemotron VL image processing (#22739) · a01e0018
  Duc-Viet Hoang authored Aug 13, 2025
```
Co-authored-by: ducviet00-h2 <viet.d.hoang@h2corporation.jp>
```
  a01e0018
- [Perf] Support topk softmax fused kernel for broader num_experts (#22211) · 4c558cf6
  shixianc authored Aug 12, 2025
```
Signed-off-by: Shixian Cui <shixian@amazon.com>
Co-authored-by: Shixian Cui <shixian@amazon.com>
```
  4c558cf6
- [V0 Deprecation] Remove args for multi-step scheduling (#22779) · c5830381
  Woosuk Kwon authored Aug 12, 2025
```
Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
```
  c5830381
- [Misc] Remove tests/multi_step/__init__.py (#22778) · d31f97cf
  Woosuk Kwon authored Aug 12, 2025
```
Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
```
  d31f97cf
- [V0 Deprecation] Remove multi-step scheduling (#22138) · 71683ca6
  Woosuk Kwon authored Aug 12, 2025
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
```
  71683ca6
12 Aug, 2025 15 commits
- [Misc] parametrize 'dtype' in test_flash_mla (#22641) · 53c73028
  RUTHLESS-BOT authored Aug 13, 2025
```
Signed-off-by: RUTHLESS-BOT <wujiafeng@cmbchina.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  53c73028
- [CI][Nixl] Check kv cache layout during handshake (#22745) · 422f22e0
  Nicolò Lucchesi authored Aug 12, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  422f22e0
- Add more test scenario for tensor schema (#22733) · c42fe0b6
  TeeKen Lau authored Aug 13, 2025
```
Signed-off-by: teekenl <teekenlau@gmail.com>
```
  c42fe0b6
- Add: `SupportsEagle3` interface for explicit EAGLE3 support (#22642) · 5a4b4b37
  Rahul Tuli authored Aug 12, 2025
```
Signed-off-by: Rahul Tuli <rtuli@redhat.com>
```
  5a4b4b37
- [Bugfix][CI] Fix `test_remote_decode_lifecycle.py::test_short_prompt_lifecycle` (#22727) · 3d9d40ef
  Nicolò Lucchesi authored Aug 12, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  3d9d40ef
- Officially support SmolLM3 using the Transformers backend (#22665) · 80bb1e8a
  Harry Mellor authored Aug 12, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  80bb1e8a
- [gpt-oss] Enable gpt-oss on ampere (#22714) · 007dd908
  Yongye Zhu authored Aug 12, 2025
```
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
```
  007dd908
- [Bugfix] Mamba2 SSD varlen bug fix initstates decay, improve test, assert chunk pwr 2 (#21783) · 46ae7f66
  RishiAstra authored Aug 12, 2025
```
Signed-off-by: Rishi Astra <40644327+RishiAstra@users.noreply.github.com>
```
  46ae7f66
- [Bugfix] Fix erroneous randomly generated cases in bad word testing (#22170) · bc8372ef
  phantomlei authored Aug 12, 2025
```
Signed-off-by: phantomlei <phantomlei3@gmail.com>
```
  bc8372ef
- [New Model] Support Command-A-Vision (#22660) · 9f909b89
  dongluw authored Aug 12, 2025
```
Signed-off-by: donglu <donglu@cohere.com>
```
  9f909b89
- [Bugfix] Fix ModernBert load & Enable sliding window attention for... · 6d729c43
  wang.yuqi authored Aug 12, 2025
```
[Bugfix] Fix ModernBert load & Enable sliding window attention for bidirectional attention. (#22637)
Signed-off-by: wang.yuqi <noooop@126.com>
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
Co-authored-by: Max de Bayser <mbayser@br.ibm.com>
```
  6d729c43
- [CI] Increase timeout for test_completion_with_image_embeds (#22670) · 93d06524
  Michael Goin authored Aug 11, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  93d06524
- [CI Failure] Use float32 for tests/entrypoints/openai/test_audio.py (#22686) · ea1292ad
  Michael Goin authored Aug 11, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  ea1292ad
- Re-enable Xet on TPU tests now that `hf_xet` has been updated (#22666) · 839ab003
  Harry Mellor authored Aug 12, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  839ab003
- [gpt-oss] Add test for response API + harmony (but skipped) (#22554) · 1891a265
  Chen Zhang authored Aug 11, 2025
```
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
```
  1891a265
11 Aug, 2025 10 commits
- [CI] Skip Tree Attn Test in `test_max_len.py` to unblock CI (#22664) · 65abe111
  TJian authored Aug 11, 2025
```
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
```
  65abe111
- [BugFix] [Spec Decode] Remove LlamaForCausalLMEagle3 to fix CI (#22611) · 807d21b8
  22quinn authored Aug 11, 2025
```
Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
```
  807d21b8
- [CI/Build] Skip Mllama HF runner tests with Transformers v4.55.0 (#22659) · c90fb03d
  Isotr0py authored Aug 12, 2025
```
Signed-off-by: Isotr0py <2037008807@qq.com>
```
  c90fb03d
- [Model] Pooling models default to using chunked prefill & prefix caching if supported. (#20930) · 84cf78ac
  wang.yuqi authored Aug 12, 2025
```
Signed-off-by: wang.yuqi <noooop@126.com>
```
  84cf78ac
- fix: NIXL connector transfers partial block to pass full multi-modal context (#21074) · 16fb668b
  GuanLuo authored Aug 11, 2025
```
Signed-off-by: GuanLuo <gluo@nvidia.com>
```
  16fb668b
- [Feature] Add `VLLM_USE_DEEP_GEMM_E8M0` Env to Control E8M0 Scale (#21968) · f7dcce7a
  Wentao Ye authored Aug 11, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  f7dcce7a
- [Misc] Move tensor schema tests (#22612) · ebf7605b
  Cyrus Leung authored Aug 11, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  ebf7605b
- Support token_type_ids in V1 with less code changes (#21985) · 39052dbc
  Maximilien de Bayser authored Aug 11, 2025
```
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
```
  39052dbc
- [BugFix] Fix KVConnectorOutput TPU breakage (#22598) · 5898b135
  Nick Hill authored Aug 10, 2025
```
Signed-off-by: Nick Hill <nhill@redhat.com>
```
  5898b135
- [CI/Build] Fix tensorizer test for load_format change (#22583) · b799f4b9
  22quinn authored Aug 10, 2025
```
Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
```
  b799f4b9
10 Aug, 2025 6 commits
- Fix TensorSchema validation test for symbolic dims (#22366) · 68b254d6
  Benji Beck authored Aug 10, 2025
```
Signed-off-by: Benji Beck <benjibeck@meta.com>
```
  68b254d6
- [Bugfix][Kernel] Support partial rotary embedding for MRoPE triton kernel (#22593) · b76753f0
  Isotr0py authored Aug 11, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  b76753f0
- [Misc] Replace flaky image urls in pixtral test (#22574) · 049c2451
  Isotr0py authored Aug 10, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: Isotr0py <2037008807@qq.com>
```
  049c2451
- [Misc] code clean duplicate set_current_vllm_config in _set_vllm_config (#22566) · 32697629
  Ning Xie authored Aug 10, 2025
```
Signed-off-by: Andy Xie <andy.xning@gmail.com>
```
  32697629
- Refactor sliding window configuration to Transformers best practice (#21927) · c4984839
  Harry Mellor authored Aug 10, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  c4984839
- [TPU] kv cache update kernel doesn't need to be padded slices to multiple of... · 2a84fb42
  Chengji Yao authored Aug 09, 2025
```
[TPU] kv cache update kernel doesn't need to be padded slices to multiple of num_slices_per_block (#22394)
Signed-off-by: Chengji Yao <chengjiyao@gmail.com>
Co-authored-by: Chengji Yao <chengjiyao@gmail.com>
```
  2a84fb42