Commits · e4755f7fac67e9dd1723173f2eb30778dbe8d792 · OpenDAS / vllm_cscc

17 Apr, 2025 9 commits

[V1][TPU] Enable Top K (#15489) · eb5819b2

Nicolò Lucchesi authored Apr 17, 2025


Signed-off-by: NickLucche <nlucches@redhat.com>
Signed-off-by: Hyesoo Yang <hyeygit@gmail.com>
Co-authored-by: Hyesoo Yang <hyeygit@gmail.com>

eb5819b2

[TPU][V1] Fix padding recompilation when `max-num-batched-tokens` is not even (#16726) · 5989f468
Nicolò Lucchesi authored Apr 17, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
5989f468
[V1][Perf] Faster incremental detokenization (#15137) · 05fcd1b4
Nick Hill authored Apr 17, 2025
```
Signed-off-by: Nick Hill <nhill@redhat.com>
```
05fcd1b4
[V1] Remove log noise when idle (#16735) · 9dbf7a2d
Russell Bryant authored Apr 17, 2025
```
Signed-off-by: Russell Bryant <rbryant@redhat.com>
```
9dbf7a2d
[Bugfix] Revert max_prompt_len validation for decoder-only models. (#16741) · 607029e5
David Heineman authored Apr 16, 2025
```
Signed-off-by: David Heineman <david@davidheineman.com>
```
607029e5

[V1][Frontend] Improve Shutdown And Logs (#11737) · 2b05b8ce

Robert Shaw authored Apr 16, 2025


Signed-off-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>
Signed-off-by: Andrew Feldman <afeldman@neuralmagic.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
Co-authored-by: Russell Bryant <rbryant@redhat.com>
Co-authored-by: Andrew Feldman <afeldman@neuralmagic.com>
Co-authored-by: afeldman-nm <156691304+afeldman-nm@users.noreply.github.com>
Co-authored-by: Nick Hill <nhill@redhat.com>

2b05b8ce

[V1][Performance] Implement custom serializaton for MultiModalKwargs [Rebased] (#16432) · 3092375e

Staszek Paśko authored Apr 17, 2025


Signed-off-by: Staszek Pasko <staszek@gmail.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
Co-authored-by: Nick Hill <nhill@redhat.com>

3092375e

[Misc] Remove redundant comment (#16703) · 8a7368e0
Jade Zheng authored Apr 17, 2025
```
Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>
```
8a7368e0
Improve error for structured output backend selection (#16717) · 93e561ec
Harry Mellor authored Apr 17, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
93e561ec

16 Apr, 2025 2 commits
- [Hardware] Add processor inputs to platform validation (#16680) · e1b00483
  Joe Runde authored Apr 16, 2025
```
Signed-off-by: Joe Runde <Joseph.Runde@ibm.com>
```
  e1b00483
- [V1][Structured Output] Move xgrammar related utils to `backend_xgrammar.py` (#16578) · 976711d9
  Shanshan Shen authored Apr 16, 2025
```
Signed-off-by: shen-shanshan <467638484@qq.com>
```
  976711d9
15 Apr, 2025 1 commit
- Add "/server_info" endpoint in api_server to retrieve the vllm_config. (#16572) · 1666e664
  Xihui Cang authored Apr 15, 2025
```
Signed-off-by: Xihui Cang <xihuicang@gmail.com>
```
  1666e664
14 Apr, 2025 4 commits
- [DOC][TPU] Add core idea about avoiding recompilation after warmup (#16614) · 1eb3c2ed
  Chengji Yao authored Apr 14, 2025
```
Signed-off-by: Chengji Yao <chengjiyao@google.com>
```
  1eb3c2ed
- [Model][VLM] Add Kimi-VL model support (#16387) · b1308b84
  courage17340 authored Apr 15, 2025
```
Signed-off-by: courage17340 <courage17340@163.com>
```
  b1308b84
- [TPU][V1] Fix exponential padding when `max-num-batched-tokens` is not a power of 2 (#16596) · b3f2fddd
  Nicolò Lucchesi authored Apr 14, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  b3f2fddd
- [Bugfix] Multi-modal caches not acting like LRU caches (#16593) · aa29841e
  Cyrus Leung authored Apr 15, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  aa29841e
13 Apr, 2025 1 commit
- [V1][Spec Decode] KV cache slots for eagle heads (#16370) · f49e5aff
  Lily Liu authored Apr 12, 2025
```
Signed-off-by: LiuXiaoxuanPKU <lilyliupku@gmail.com>
```
  f49e5aff
12 Apr, 2025 4 commits
- [Bugfix] Validate logit biases to prevent out of vocab ids crashing engine (#16529) · 6c11ecf8
  Ryan McConville authored Apr 12, 2025
```
Signed-off-by: Ryan McConville <ryan@ryanmcconville.com>
```
  6c11ecf8
- [Perf] Optimize Preparing Inputs for GPU Model Runner (#16484) · 93e5f3c5
  SnowCharm authored Apr 12, 2025
```
Signed-off-by: snowcharm <snowcharmqq@gmail.com>
Co-authored-by: Nick Hill <nhill@redhat.com>
```
  93e5f3c5
- [Feature][V1] Add xgrammar to support minLength, maxLength with test (#16516) · e92d7085
  leon-seidel authored Apr 12, 2025
```
Signed-off-by: Leon Seidel <leon.seidel@fau.de>
```
  e92d7085
- [BugFix] Handle non-contiguous tensors properly when serializing (#16492) · 41cc883c
  Nick Hill authored Apr 11, 2025
```
Signed-off-by: Nick Hill <nhill@redhat.com>
```
  41cc883c
11 Apr, 2025 2 commits
- [Kernel] support merge_attn_states CUDA kernel, 3x speedup (#16173) · e9528f6d
  DefTruth authored Apr 11, 2025
```
Signed-off-by: DefTruth <qiustudent_r@163.com>
```
  e9528f6d
- Enforce valid max_num_batched_tokens when disable_chunked_mm_input=True (#16447) · aa3b3d76
  Michael Goin authored Apr 11, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  aa3b3d76
10 Apr, 2025 6 commits
- [TPU][V1] Disable per-request seed/Generator (#16172) · 3cc9af88
  Nicolò Lucchesi authored Apr 10, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  3cc9af88
- [V1] Zero-copy tensor/ndarray serialization/transmission (#13790) · dd143ef5
  Nick Hill authored Apr 10, 2025
```
Signed-off-by: Nick Hill <nhill@redhat.com>
```
  dd143ef5
- [V1][Spec Decode] Eagle Model loading (#16035) · e8224f3d
  Lily Liu authored Apr 10, 2025
```
Signed-off-by: LiuXiaoxuanPKU <lilyliupku@gmail.com>
```
  e8224f3d
- [TPU][V1] Use `language_model` interface for getting text backbone in MM (#16410) · c1b57855
  Nicolò Lucchesi authored Apr 10, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  c1b57855
- [Bugfix] Fix validation error for text-only Mllama 3.2 (#16377) · a5d11a54
  Cyrus Leung authored Apr 10, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  a5d11a54
- [TPU][V1] Refine tpu_model_runner to mitigate future recompilation issues (#16275) · a4547485
  Chengji Yao authored Apr 09, 2025
```
Signed-off-by: Chengji Yao <chengjiyao@google.com>
```
  a4547485
09 Apr, 2025 5 commits
- [Hardware] add platform-specific request validation api (#16291) · cb391d85
  Joe Runde authored Apr 09, 2025
```
Signed-off-by: Joe Runde <Joseph.Runde@ibm.com>
```
  cb391d85
- [BugFix] fix some typos found by typos. (#16314) · 04149cce
  yihong authored Apr 09, 2025
```
Signed-off-by: yihong0618 <zouzou0208@gmail.com>
```
  04149cce
- [Bugfix] Avoid transferring cached multi-modal items from P0 to P1 (#16273) · e484e028
  Cyrus Leung authored Apr 09, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  e484e028
- [Core] Upgrade to xgrammar 0.1.18, add cache size limit (#16283) · cb84e45a
  Russell Bryant authored Apr 08, 2025
```
Signed-off-by: Russell Bryant <rbryant@redhat.com>
```
  cb84e45a
- [Feature] Estimate max-model-len use available KV cache memory (#16168) · 4716377f
  rongfu.leng authored Apr 09, 2025
```
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
```
  4716377f
08 Apr, 2025 6 commits
- [BugFix] Fix Llama4 - Index Error When Single Request Near Max Context (#16209) · e1a2c699
  Lucas Wilkinson authored Apr 08, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  e1a2c699
- [Bugfix] Proper input validation for multi-modal encoder-decoder models (#16156) · 4ebc0b96
  Cyrus Leung authored Apr 09, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  4ebc0b96
- [V1] Add `disable_chunked_mm_input` arg to disable partial mm input prefill (#15837) · 8e5314a4
  Michael Goin authored Apr 08, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  8e5314a4
- [torch.compile][TPU] Make @support_torch_compile work for XLA backend (#15782) · 87918e40
  Siyuan Liu authored Apr 07, 2025
```
Signed-off-by: Siyuan Liu <lsiyuan@google.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
```
  87918e40
- Add warning for Attention backends that do not support irope yet (#16212) · 05a015d6
  Yong Hoon Shin authored Apr 07, 2025
  
  05a015d6
- [V1] Scatter and gather placeholders in the model runner (#16076) · f2ebb6f5
  Roger Wang authored Apr 07, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Roger Wang <ywang@roblox.com>
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
Co-authored-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Jennifer Zhao <ai.jenniferzhao@gmail.com>
```
  f2ebb6f5