Commits · b18201fe060a3ddcc088f8aea3cf1d7c4b461288 · OpenDAS / vllm_cscc · GitLab

16 May, 2025 3 commits
- Allow users to pass arbitrary JSON keys from CLI (#18208) · b18201fe
  Harry Mellor authored May 16, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  b18201fe
- [Model] vLLM v1 supports Medusa (#17956) · f4937a51
  Sky Lee authored May 16, 2025
```
Signed-off-by: lisiqi23 <lisiqi23@xiaomi.com>
Signed-off-by: skylee-01 <497627264@qq.com>
Co-authored-by: lisiqi23 <lisiqi23@xiaomi.com>
```
  f4937a51
- [Bugfix][ROCm] Use `chunked_prefill_paged_decode` as fallback for V1 attention on ROCm (#18093) · ee659e3b
  kliuae authored May 16, 2025
```
Signed-off-by: kf <kuanfu.liu@embeddedllm.com>
```
  ee659e3b
15 May, 2025 18 commits
- [Misc] Avoid cuda graph log when sizes still match (#18202) · e3f3aee6
  Nicolò Lucchesi authored May 15, 2025
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  e3f3aee6
- [Bugfix] [ROCm]: Remove assertion logic when using AITER fused moe in... · 92540529
  TJian authored May 16, 2025
```
[Bugfix] [ROCm]: Remove assertion logic when using AITER fused moe in unquantizedMethod to reenable LLama4 BF16 (#18205)
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
```
  92540529
- [Bugfix]Change the exception thrown by call_hf_processor from RuntimeError to ValueError (#18181) · fadb8d5c
  Zhonghua Deng authored May 16, 2025
```
Signed-off-by: Abatom <abzhonghua@gmail.com>
```
  fadb8d5c
- [Frontend] Fix chat template content format detection (#18190) · 2aa5470a
  Sebastian Schoennenbeck authored May 15, 2025
```
Signed-off-by: Sebastian Schönnenbeck <sebastian.schoennenbeck@comma-soft.com>
```
  2aa5470a
- [Kernel] [V1] Fix performance regression for triton unified attention (#18161) · 01c22335
  Thomas Parnell authored May 15, 2025
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  01c22335
- add tools into TokenizeChatRequest (#18187) · 451da4bc
  hustxiayang authored May 15, 2025
```
Signed-off-by: yangxia <yangxiast@gmail.com>
```
  451da4bc
- Update deprecated type hinting in `model_loader` (#18130) · 07ad2712
  Harry Mellor authored May 15, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  07ad2712
- fix: typos (#18151) · a9944aab
  omahs authored May 15, 2025
```
Signed-off-by: omahs <73983677+omahs@users.noreply.github.com>
```
  a9944aab
- [V1] Update zmq socket creation in nixl connector (#18148) · a8f5aec2
  Russell Bryant authored May 15, 2025
```
Signed-off-by: Russell Bryant <rbryant@redhat.com>
```
  a8f5aec2
- [Bugfix] Fix FusedMoEPrepareAndFinalize for cuda-disalike backends (#18178) · 70f8b967
  Mengqing Cao authored May 15, 2025
```
Signed-off-by: Mengqing Cao <cmq0113@163.com>
```
  70f8b967
- [Model] Allow the use of sliding window in Qwen2 (#17772) · dd2a9459
  inkcherry authored May 15, 2025
```
Signed-off-by: inkcherry <mingzhi.liu@intel.com>
```
  dd2a9459
- Support custom implementations of VideoLoader backends. (#18091) · 4f07a640
  Chenheli Hua authored May 14, 2025
  
  4f07a640
- [Bugfix] Fix fp8 tests for triton_unified_attention for Triton 3.3 (#18013) · e6b8e65d
  Thomas Parnell authored May 15, 2025
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  e6b8e65d
- Update deprecated type hinting in `models` (#18132) · 26d04193
  Harry Mellor authored May 15, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  26d04193
- [Fix][ROCm] Enforce eager for all encoder-decoder models on ROCm (#18154) · 83f74c69
  Luka Govedič authored May 15, 2025
```
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
```
  83f74c69
- [V1][Metrics] Remove unused code (#18158) · 65334ef3
  Mark McLoughlin authored May 15, 2025
```
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
```
  65334ef3
- [v1] Support multiple KV cache groups in GPU model runner (#17945) · e60f550b
  Chen Zhang authored May 15, 2025
```
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
```
  e60f550b
- [Bugfix]: make most of `test_openai_schema.py` pass (#17664) · f25e0d11
  David Xia authored May 14, 2025
  
  f25e0d11
14 May, 2025 19 commits
- [V1] Support multiple kv connectors (#17564) · 2142035b
  Michael Goin authored May 14, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: Nick Hill <nhill@redhat.com>
```
  2142035b
- [CI] Fix race condition in test_kv_cache_events test (#18169) · 78aa341d
  Russell Bryant authored May 14, 2025
```
Signed-off-by: Russell Bryant <rbryant@redhat.com>
```
  78aa341d
- Add support for loading torchao models with `AOPerModuleConfig` (#17826) · 79747367
  Jerry Zhang authored May 14, 2025
```
Signed-off-by: Jerry Zhang <jerryzh168@gmail.com>
```
  79747367
- [V1] Structured Outputs + Thinking compatibility (#16577) · 2fc9075b
  Aaron Pham authored May 14, 2025
```
Signed-off-by: Aaron Pham <contact@aarnphm.xyz>
Co-authored-by: Russell Bryant <rbryant@redhat.com>
```
  2fc9075b
- [Kernel] Have rotary embeddings support tensors (#18046) · d93c976a
  Lucas Wilkinson authored May 14, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  d93c976a
- [Frontend] decrease import time of vllm.multimodal (#18031) · 749f7925
  David Xia authored May 14, 2025
```
Co-authored-by: Aaron Pham <Aaronpham0103@gmail.com>
```
  749f7925
- Modularize fused experts and integrate PPLX kernels (#15956) · f9c069c8
  bnellnm authored May 14, 2025
  
  f9c069c8
- [V1][Spec Decode] Share input embedding of target model with EAGLE draft model... · 418d2f8b
  Ekagra Ranjan authored May 14, 2025
```
[V1][Spec Decode] Share input embedding of target model with EAGLE draft model to free ~1GB for llama 3 model (#17326)
Co-authored-by: root <root@ekagra-8xh100.us-east5-a.c.serving-efficiency-poc.internal>
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  418d2f8b
- [KVConnector] Keep KVTransferParams as a dict (#18033) · 59dd311c
  Nick Hill authored May 14, 2025
  
  59dd311c
- Update deprecated type hinting in `platform`, `plugins`, `triton_utils`, `vllm_flash_attn` (#18129) · c8ea982d
  Harry Mellor authored May 14, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  c8ea982d
- Update deprecated type hinting in `vllm/device_allocator` and `vllm/distributed` (#18126) · dc372b9c
  Harry Mellor authored May 14, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  dc372b9c
- Update deprecated type hinting in `vllm/lora` (#18128) · 9b5b39b6
  Harry Mellor authored May 14, 2025
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  9b5b39b6
- [Model] GritLM supports other attention backends (#18109) · d62a076e
  Cyrus Leung authored May 14, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  d62a076e
- [FEAT] [ROCm]: Add AITER CK 2 Stages MoE support (#17110) · 612c2edb
  TJian authored May 14, 2025
```
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
Co-authored-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
```
  612c2edb
- [Bugfix] Fix QKVCrossParallelLinear::sync_weight_attrs for PyTorch compile (#17844) · 38fe728d
  Andrzej Kotłowski authored May 14, 2025
```
Signed-off-by: Andrzej Kotłowski <akotlowski@habana.ai>
```
  38fe728d
- [Misc] replace does not exist model (#18119) · 82e7f9bb
  rongfu.leng authored May 14, 2025
```
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
```
  82e7f9bb
- [Model] Add packed_modules_mapping for Qwen3-MOE (#18118) · 63dc3426
  Jee Jee Li authored May 14, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  63dc3426
- [New Model]: support GTE NewModel (#17986) · 63ad6222
  wang.yuqi authored May 14, 2025
  
  63ad6222
- [Fix] Move "model_config" as keyword args in chat_utils.py (#18098) · 6685890d
  lkchen authored May 13, 2025
```
Signed-off-by: Linkun <github@lkchen.net>
```
  6685890d