anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity
1,110 Commits 235 Branches 31 Tags
update-committers
Commit Graph
699 Commits
Author SHA1 Message Date
R0CKSTAR 3def008b32 [Feat] Add MUSA platform support for Moore Threads GPUs (#2337) 
Signed-off-by: Xiaodong Ye <yeahdongcn@gmail.com>
 2026-04-01 17:55:00 +08:00
汪志鹏 bf5bd0a4c0 [BugFix]: Fix bagel single-stage img2img fallback to text2img bug (#2397) 
Signed-off-by: princepride <wangzhipeng628@gmail.com>
 2026-04-01 02:29:02 -07:00
Wu JIAZHENandjader 3fd4a4dc27 [Feat] Support step-boundary abort in diffusion (#1769) 
Signed-off-by: jader <yjader@foxmail.com>
Signed-off-by: asukaqaq-s <1311722138@qq.com>
Co-authored-by: jader <yjader@foxmail.com>
 2026-04-01 17:25:02 +08:00
NATURE d40840b114 [Perf] Bagel KV-ready early forwarding and time step consistency for /v1/chat/completions (#2398) 
Signed-off-by: natureofnature <wzliu@connect.hku.hk>
 2026-04-01 08:50:45 +00:00
zdobaanddaixinning 08cb436d4e Fix: ensure input tensor is contiguous in GroupCoordinator.all_gather (#2367) 
Signed-off-by: daixinning <daixinning@163.com>
Co-authored-by: daixinning <daixinning@163.com>
 2026-04-01 16:41:25 +08:00
Ding Zuhao c3376a466b [Feature][HunyuanImage3.0] Add cfgP to HunyuanImage3.0 (#1751) 
Signed-off-by: Ding Zuhao <e1583181@u.nus.edu>
 2026-04-01 16:33:02 +08:00
Samit 7b965a7b6c [Feat][Benchmark] Add synchronous video generation endpoint POST /v1/videos/sync for benchmark test (#2049) 
Signed-off-by: samithuang <285365963@qq.com>
 2026-04-01 10:47:33 +08:00
fattysand 7274e15840 [BugFix] qwen3_tts chunk boundary handling logic in initial chunk (IC) (#2378) 
Signed-off-by: Fattysand <fattysand@users.noreply.github.com>
 2026-04-01 09:40:12 +08:00
Lancer dd0b6fd4e8 [Bugfix] Fix layer-wise offload incompatibility with HSDP (#2021) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-04-01 09:25:34 +08:00
汪志鹏 369f301a1b [BugFix] Add BAGEL single-stage diffusion config and fix multiple <im_start><im_end> bug (#2381) 
Signed-off-by: princepride <wangzhipeng628@gmail.com>
 2026-04-01 09:19:13 +08:00
Reid de1ac5ee88 [Bugfix] Fix case-sensitivity in Qwen3 TTS speaker name lookup (#2358) 
Signed-off-by: reidliu41 <reid201711@gmail.com>
 2026-03-31 09:51:35 +00:00
LLandCanlin Guo 12a8f57934 Fix Qwen3-TTS Base on NPU running failed (#2353) 
Signed-off-by: LL <rimchew@foxmail.com>
Signed-off-by: Canlin Guo <canlinguosdu@gmail.com>
Co-authored-by: Canlin Guo <canlinguosdu@gmail.com>
 2026-03-31 12:42:21 +08:00
Yangshen DengandClaude Opus 4.6 837679dbaa [Diffusion] Refactor CFG parallel for extensibility and performance (#2063) 
Signed-off-by: Yangshen Deng <yangshen.d@outlook.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-30 16:48:58 +00:00
WeiQing Chen 1ca942999a [Bugfix] fix: validate layered image layers range (#2334) 
Signed-off-by: David Chen <530634352@qq.com>
 2026-03-30 21:27:09 +08:00
WeiQing Chen 1eee8435e1 [Bugfix] fix: return 400 for unsupported multi-image edits such as Qwen-Image-Layered (#2298) 
Signed-off-by: David Chen <530634352@qq.com>
 2026-03-30 20:24:57 +08:00
wangyuyenuo26R2-YYueqian Lin
 		c9370f493e [Test] Add Qwen-tts test cases and unify the style of existing test cases (#2195) 
Signed-off-by: yenuo26 <410167048@qq.com>
Signed-off-by: wangyu <53896905+yenuo26@users.noreply.github.com>
Signed-off-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
Signed-off-by: wangyu <410167048@qq.com>
Co-authored-by: yenuo26 <410167048@example.com>
Co-authored-by: R2-Y <ruiruyang2@gmail.com>
Co-authored-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
 2026-03-30 01:02:28 -04:00
Lancer 32af3af262 [Bugfix] Fix Qwen3TTSConfig init order to be compatible with newer Tansformers(5.x) (#2306) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-03-30 09:50:16 +08:00
WeiQing ChenandCanlin Guo ecfee25c3b [Bugfix] fix: handle Qwen-Image-Layered layered RGBA output for jpeg edits (#2297) 
Signed-off-by: David Chen <530634352@qq.com>
Co-authored-by: Canlin Guo <canlinguosdu@gmail.com>
 2026-03-29 11:33:37 +00:00
rongfu.lengandHongsheng Liu 9d9e6abae6 [Feat] add helios fp8 quantization (#1916) 
Signed-off-by: rongfu.leng <lenronfu@gmail.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-03-29 06:35:56 +00:00
Ding Zuhaoand汪志鹏 283a62fd76 [Feature] [HunyuanImage3] Add TeaCache support for inference acceleration (#1927) 
Signed-off-by: Ding Zuhao <e1583181@u.nus.edu>
Signed-off-by: 汪志鹏 <wangzhipeng628@gmail.com>
Co-authored-by: 汪志鹏 <wangzhipeng628@gmail.com>
 2026-03-28 15:24:21 +00:00
Junhong Liu a3e1322ad9 Qwen3-Omni][Bugfix] Replace vLLM fused layers with HF-compatible numerics in code predictor (#2291) 
Signed-off-by: Junhong Liu <98734602+LJH-LBJ@users.noreply.github.com>
 2026-03-28 23:02:44 +08:00
lif 1414195fbb Fix: Enable /v1/models endpoint for pure diffusion mode (#805) 
Signed-off-by: majiayu000 <1835304752@qq.com>
 2026-03-28 22:17:38 +08:00
Lancer 64629cc174 [Feat] support quantization for Flux Kontext (#2184) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-03-28 10:28:10 +00:00
Ding Zuhao 5cf6870277 [Bugfix][HunyuanImage3.0] Fix default guidance_scale from 1.0 to 4.0 and port GPU MoE ForwardContext fix from NPU (#2142) 
Signed-off-by: Ding Zuhao <e1583181@u.nus.edu>
 2026-03-28 13:49:03 +08:00
Yueqian Lin f55ea28005 [Qwen3TTS][Bugfix] Replace vLLM fused layers with HF-compatible numerics in code predictor (#2277) 
Signed-off-by: linyueqian <linyueqian@outlook.com>
 2026-03-28 10:47:17 +08:00
Sy03 d9912eb6e9 [Bugfix] Fix Fish Speech S2 Pro prompt handling for truncated audio & emotion tag (#2268) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-03-28 07:00:00 +08:00
dengyunyangand汪志鹏 c0d90b64c6 [Feature] support sp for hunyuan (#2163) 
Signed-off-by: dengyunyang <584797741@qq.com>
Signed-off-by: 汪志鹏 <wangzhipeng628@gmail.com>
Co-authored-by: 汪志鹏 <wangzhipeng628@gmail.com>
 2026-03-27 17:28:13 +08:00
Junhong Liu c51adeacde [CI] Add online e2e test for MIMO-Audio (#2129) 
Signed-off-by: Junhong Liu <98734602+LJH-LBJ@users.noreply.github.com>
 2026-03-27 17:19:54 +08:00
knlnguyen1802andSamit 49c1abea17 [Bugfix] Fix dynamic function call on collective_rpc of DiffusionWorker (#2217) 
Signed-off-by: knlnguyen1802 <knlnguyen1802@gmail.com>
Co-authored-by: Samit <285365963@qq.com>
 2026-03-27 14:28:59 +08:00
amy-why-3459 9f667ee151 [BugFix][Qwen3-Omni]Fixed the issue of incorrect answers for single words. (#2239) 
Signed-off-by: amy-why-3459 <wuhaiyan17@huawei.com>
 2026-03-27 11:46:24 +08:00
Lancer 055963b3fd [Bugfix] Fix tp and Quantization incompatible for Flux (#2180) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-03-27 09:21:50 +08:00
Cheung Ka Wai 590ab5b3e9 [Bug Fix] Resolve broken image issue when TP is enabled and no seed is provided. (#2176) 
Signed-off-by: Mike Cheung <zhtmike@gmail.com>
 2026-03-26 16:56:46 +00:00
Baoyuan Qi 66d5478dad [Compatibility] Add Multiple Attention Backends Support in MIMO-Audio Tokenizer (#2225) 
Signed-off-by: 齐保元 <qibaoyuan@xiaomi.com>
 2026-03-27 00:47:53 +08:00
Juan Pablo ZuluagaandYueqian Lin 574ec99ef4 [Qwen3TTS][ServingSpeech] Bugfix/voice upload and add optional ref_text (#2046) 
Signed-off-by: JuanPZuluaga <juanz9312@gmail.com>
Co-authored-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
 2026-03-26 14:26:52 +00:00
Yueqian Lin 9857ed2839 [Bugfix] Fix Voxtral TTS voice embeddings not loading (#2156) 
Signed-off-by: linyueqian <linyueqian@outlook.com>
 2026-03-26 21:34:31 +08:00
Jared Wen 15df8ec97f [Bug Fix] GLM-Image stage device isolation and t2i prompt preprocessing in Omni runtime (#2137) 2026-03-26 10:50:54 +00:00
Daniel Huang ce916f4233 [Enhancement] Patch AsyncOmniEngine try_get_output[_async] hanging issues (#2153) 
Signed-off-by: Daniel Huang <daniel1.huang@intel.com>
 2026-03-26 16:52:01 +08:00
amy-why-3459 029c85393a [BugFix]Fix keyError: num_processed_tokens_delta (#2213) 
Signed-off-by: amy-why-3459 <wuhaiyan17@huawei.com>
 2026-03-26 15:16:25 +08:00
Sy03 d15c91d6d2 [Bugfix] Fix qwen3-omni async thinker to talker decode alignment for #1758 (#2019) 2026-03-26 11:07:11 +08:00
Yuanheng Zhao 04a6e5a992 [Bugfix] Maintain model-level CPU offload in a blocking way (#1978) 
Signed-off-by: Yuanheng Zhao <jonathan.zhaoyh@gmail.com>
 2026-03-26 02:08:13 +00:00
Lancer 9f1d61b0f7 [Bugfix] resolve stage config for GLM-Image with diffusers format (#1894) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-03-26 09:47:01 +08:00
rongfu.leng b569cb56b6 [Bugfix] add inject model_arch to hf_overrides (#2178) 
Signed-off-by: rongfu.leng <lenronfu@gmail.com>
 2026-03-25 19:05:28 +00:00
Markus / MarkYueqian Linlinyueqian
 		29936a3edd [Frontend] Speaker embedding support for speech and voices APIs (#1227) 
Signed-off-by: marksverdhei <marksverdhei@hotmail.com>
Signed-off-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
Signed-off-by: linyueqian <linyueqian@outlook.com>
Co-authored-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
Co-authored-by: linyueqian <linyueqian@outlook.com>
 2026-03-25 19:02:48 +00:00
Sy03 fc32da7624 [Perf] [TTS] Improve Fish Speech S2 Pro voice cloning TTFP (#2145) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-03-25 11:41:10 -04:00
Alex BrooksandSYLAR c09b039f37 [BugFix] Make Stage Device Initialization Respect Visible GPUs (#2025) 
Signed-off-by: Alex Brooks <albrooks@redhat.com>
Co-authored-by: SYLAR <125541396+lishunyang12@users.noreply.github.com>
 2026-03-25 17:56:13 +08:00
Hui.andAlvisGong 0adbac33a5 [Bug-Fix]fix bug of empty prompt input (#2041) 
Signed-off-by: AlvisGong <gwly0401@163.com>
Signed-off-by: Hui <1779066624@qq.com>
Signed-off-by: Hui. <1779066624@qq.com>
Co-authored-by: AlvisGong <gwly0401@163.com>
 2026-03-25 16:42:04 +08:00
Yueqian LinandCanlin Guo 91a818629c [Cleanup] Remove stray test file from engine directory (#2161) 
Signed-off-by: linyueqian <linyueqian@outlook.com>
Co-authored-by: Canlin Guo <canlinguosdu@gmail.com>
 2026-03-25 08:26:38 +00:00
Junhong Liu f3d078e9f2 [CI] Add online e2e test for qwen2.5 omni (#1668) 
Signed-off-by: Junhong Liu <98734602+LJH-LBJ@users.noreply.github.com>
Signed-off-by: Junhong Liu <ljh_lbj@163.com>
 2026-03-25 14:14:34 +08:00
Chen-Yo Sun e776082095 [Voxtral TTS] Use 8 step flow matching instead of 16 (#2158) 
Signed-off-by: Chen-Yo Sun <chenyo.sun@mistral.ai>
 2026-03-25 04:33:18 +00:00
Sy03 44770f9f20 [Fix] Qwen3 TTS audio handling for long ref_audio (#2104) 
Signed-off-by: Sy03 <1370724210@qq.com>
Signed-off-by: Sy0307 <1370724210@qq.com>
 2026-03-24 20:02:52 +00:00